Coupling Impedance of the Cern Sps Beam Position Monitors

A detailed knowledge of the beam coupling impedance of the CERN Super Proton Synchrotron (SPS) is required in order to operate this machine with a higher intensity for the foreseen Large Hadron Collider (LHC) luminosity upgrade. A large number of Beam Position Monitors (BPMs) is currently installed in the SPS, and this is why their contribution to the SPS impedance has to be assessed. This paper focuses on electromagnetic (EM) simulations and bench measurements of the longitudinal and transverse impedance generated by the horizontal and vertical BPMs installed in the SPS machine. INTRODUCTION Machine studies performed at the CERN SPS since 2002 have shown that single-bunch proton beams with low longitudinal emittance are affected by heavy losses after less than one synchrotron period following injection as a result of a fast vertical instability [1,2]. These observations have triggered a detailed study of the impedance of the main elements of the SPS machine in order to complete the work already started in the longitudinal plane at the end of the nineties [3] and to create a database of the longitudinal and transverse impedances of the elements of the SPS machine. The results of electromagnetic field simulations of the Horizontal (BPH) and Vertical (BPV) beam position monitors are discussed in the next paragraphs. These elements have been chosen because of their large number (106 BPH and 96 BPV in 2006’s run). The shape and materials of these SPS BPMs are not trivial and to our knowledge there is no theory to predict their beam coupling impedance. As a consequence, bench RF measurements were also performed to benchmark the simulation results with available measured observables. EM SIMULATIONS SETUP The geometric structures of the BPH and BPV were first generated with the first version of MAFIA [4], but all geometries and simulation results presented here were obtained with CST Studio Suite [5]. Pictures of an actual BPH are shown in Fig. 1, and 3D representations of an SPS BPH used for simulations in Fig. 2. Many simplifications were applied to obtain the modelled geometry used for the simulations: (1) the outer cylindrical shell and the cavity between this shell and the rectangular inner body observed in Fig. 1 (a) were not modelled, thereby assuming the fields created by the beam can not reach beyond the inner body. (2) Several features of the inner casing were removed (electrodes’ screws, brass calibration plates, ceramic plots) or simplified (perfectly matched electrode coaxial port, casing to beam pipe transition, shape of the beam pipe cross-section, mechanical tolerances). (3) The SPS BPM inner body is in Anticorodal B [6], and both the electrodes and the beam pipe are in Stainless Steel 304L. However, all metallic parts of the simulated BPMs were assumed to be made of perfect conductors (PEC). These simplifications are assumed to have a smaller impact on the results than the coupling between the cavities behind the electrodes and the main cavity through which the beam traverses. Figure 1: Pictures of an SPS BPH: full BPH assembly (a), and partly dismantled BPH inner body (b). CST Particle Studio’s Wakefield Solver was used to obtain the wake potential generated by the passage of a gaussian bunch through the BPM. The Beam Coupling Impedance of the BPM is also automatically postprocessed from this wake potential. The Eigenmode solver of CST Microwave Studio was used to obtain the parameters of the modes trapped in the structure (resonance frequency fres, shunt impedance Rs and quality factor Q). Figure 2: Model of the SPS BPH: (a) Beam path and wake integration path are shown with blue and orange arrows. Using a cut plane (b) reveals the inner structure of the BPH body: a pair of bi-triangular shaped electrodes isolated from the rest of the casing with vacuum cavities. Beam entry and exit planes are perfect matching layers, and all other boundaries are PEC. TIME DOMAIN SIMULATIONS For the following CST Particle Studio simulations, the 15 meter wake potential left by a 1 cm rms bunch in a BPH model made of 1 million mesh cells was calculated using the indirect testbeam wakefield solver. The vertical impedance of the SPS is more critical, and this why only results for the vertical impedance are shown in Fig. 3. The (a) (b) FR5RFP049 Proceedings of PAC09, Vancouver, BC, Canada 4646 Beam Dynamics and Electromagnetic Fields D04 Instabilities Processes, Impedances, Countermeasures 0 1 2 3 4 -500 0 500 Lo ng itu di na l i m pe da nc e (in Ω ) Frequency (in GHz) Real Part