Experimental studies on the SPS electron cloud

Abstract

One of the most important limitations in the performances of the CERN-SPS is presently the Electron Cloud Instability (ECI). Hence, defining its dependence on energy with confidence is an indispensable asset to direct the efforts for all the upgrade studies. Macroparticle simulations carried out with the HEADTAIL code [1] have shown that the ECI mechanism is subtle and the scaling laws valid for the Transverse Mode Coupling Instability cannot be applied to it [2]. The reason lies in the fact that the electron dynamics, while a bunch is going through an electron cloud, is heavily affected by the transverse beam size. In fact, transversely smaller beams can enhance the electron pinch and lower the intensity threshold for the bunch to be unstable. Hence, higher energy beams, though more rigid, can be more unstable due to their smaller transverse size (with constant transverse normalized emittance). During the 2007 run a measurement campaign has been carried out at the CERN-SPS to prove experimentally the outcomes of macroparticle simulations. INTRODUCTION AND MOTIVATIONS Plans for the Large Hadron Collider (LHC) performance upgrade include the improvement of the existing LHC injectors and/or the design of possible new rings in the injector chain [3]. Several scenarios, aimed at overcoming the existing bottlenecks, are presently being taken into consideration. One option, based on the replacement of the Proton Synchrotron (PS) ring with the PS2 [4], foresees an increase of the injection energy into the existing SPS from the present 26 GeV/c to 50 GeV/c. This is believed to be beneficial for the machine in many regards (e.g., less space charge and intra beam scattering, more rigid beams against coupled bunch instabilities, no transition crossing, lower injection and capture losses) [2]. Furthermore, it would allow for an upgrade of the SPS to a 1 TeV extraction energy ring, with the related advantages for injection into the LHC. However, the SPS upgrade plan crucially depends on the effect of a higher injection energy on the collective phenomena that are presently believed to be the real limitation in the SPS performance. One of them is TMCI, which was observed in the SPS for special intense bunches with low longitudinal emittance [5, 6]. Therefore, it could be a potential limiting factor in the future, especially taking ∗Giovanni.Rumolo@cern.ch into account the enhancement of the impedance of the SPS caused by the installation of 9 new extraction kickers in the ring since 2003 and the higher charge per bunch that should be injected into the SPS [7]. In addition, the vertical single bunch ECI has been limiting for a long time the number of batches that could be injected into the SPS and it could be overcome by beam scrubbing and subsequently operating the ring with a high vertical chromaticity (which nonetheless can be harmful for the beam lifetime) [8]. A detailed study on the energy dependence of the threshold for the onset of these instabilities is essential to assess a global beneficial effect of the pre-injector upgrade without unwanted side effects. The scaling law of the TMCI threshold with energy was already addressed in [9]. Under conservation of the longitudinal emittance and assuming bunches always matched to their buckets, the TMCI threshold only depends linearly on the slip factor |η|, and therefore a higher injection energy would certainly help to operate the machine farther from this limitation. Besides, preliminary studies of the dependence of the ECI threshold on energy were done, which showed that the related scaling law cannot be trivially derived from the existing TMCI theories. In fact, a first attempt of analytical approach using a broad-band resonator with beam dependent parameters showed that it may become surprisingly unfavourable at high energies far from transition, under the further assumptions of conservation of the bunch length and the normalized transverse emittances. A comprehensive study of the effect of higher injection energy on the ECI has been therefore carried out numerically and experiments are being done in the CERN-SPS with an LHC-type beam to verify it. SUMMARY OF SIMULATION RESULTS AND CODE-TO-CODE BENCHMARK Table 1 shows a list of the essential parameters used for the numerical study (typical LHC-type bunch in the SPS). The main assumptions of our model are: • The longitudinal emittance and the bunch length are kept constant. The momentum spread ∆p/p0 is re-scaled and the matched voltage re-adjusted accordingly when changing the energy. The matched voltage goes like |η|/γ with energy. This constraint could be relaxed by increasing the longitudinal emittance. • The normalised transverse emittances are constant. BEAM’07 PROCEEDINGS