----------------------------------------------------------- Minutes of the ABP-RLC team meeting of 17.02.2006 present: RA, FC, UD, AG, WH, EM, TP, FR, GR, FS, ES, RT, FZ, Jan Koopman, Federico Roncarolo, Thomas Weiler web site: http://ab-abp-rlc.web.cern.ch/ab-abp-rlc/ ------------------------------------------------------------ (1) Minutes of last meeting, pending actions ------------------------------------------------------------ TDI impedance measurements are being pursued by FC et al. (2) Update on bb impedance of TCTV collimator and follow-up (AG) ---------------------------------------------------------------- AG recalled the geometry of the TCTV and TCLIA collimators. Refining the mesh size in GdfidL doubled the broadband impedance from 300 kOhm/m, which was already unacceptable, to 600 kOhm/m per device. Also performing a more precise "Napoly integration", the code still gives the same high value. Changing the taper angle from 15 degrees to 7 degrees or using a nonlinear taper makes no difference, which was unexpected, but it seems to suggest that the broadband impedance does not arise from the taper or that there is a bug in GdfidL. Indeed the broadband impedance is 10-20 times higher than that predicted by an analytical formula (Yokoya) for a simple taper. AG studied the simple taper geometry with GdfidL and here obtained an impedance of 35 kOhm/m, which is consistent with the analytical estimate. The input format for the simple geometry was different from that for the TCTV, however. The latter was converted from HFSS, which might be a source of problems. Also for this simple geometry, it was found that a finer mesh increased the broadband impedance by about a factor of two (see slides). FR recommended assuming that the GdfidL calculation is correct. In this case, the gap will need to be increased in order to obtain an acceptable impedance. The team was reminded that the trapped modes were also significant (see previous meeting). AG recommended the removal of the sliding rf contacts and the installation of ferrites on the outer side. => ACTION => Inform Manfred Meyer of TCTV changes suggested (AG) DONE after the meeting. AG sent the following information (shortened): "TRAPPED MODES: To reduce impedance of dipole trapped modes we need to remove the sliding rf fingers along the jaw sides which provide an rf contact between the jaws and the beam screen. This will couple the trapped modes to the external vacuum tank where they can be damped using ferrites. The ferrite material has to be placed outside of beam screen on the inner walls of the vacuum tank. Ferrite material has to form a ring around beam screen on both ends of the collimator vacuum tank. TAPER: We did not see big influence of taper angle and shape on the BB impedance of the TCTV but we think it is in any case a good idea to keep it as small as possible. We have no other options to improve rf performance of TCTV at given gap of 3mm." => ACTION => Inform Warner Bruns of suspected GdfifL bug. DONE. AG sent an email in the afternoon of the same day. The beta function needs to be folded in when making estimates of beam instability. They were included in EM's calculations, but possibly not when comparing the broadband impedance. The present gap size was determined from global requirements taking an IP beta function of 0.55 m at IP1 and 5. RA confirmed that the gap could be increased until the limit required by machine protection is reached. => ACTION => Compute collimation performance and loss maps with 2 m beta-star at IP2 and IP8 and larger TCTV gaps (RA,TW) RA mentioned that even smaller collimator gaps will be required for operating the Roman Pots. ES suggested to check the heating of the ferrite. FZ proposed to include Warner Bruns in the RLC-meeting invitation list, for questions concerning GdfidL. (3) Impedance measurements of LHC wire scanners (FC) ---------------------------------------------------- TK and FC used the wire technique to measure the impedance of the wire scanners. Each wire scanner has two parking positions. The wires are made from carbon with a dc resistance of 2.2 kOhm. They act as pick ups. At high frequencies (>1 GHz) adding ferrites suppresses the power measured in the outer wire cavities. A strong resonance at 180 MHz in the parking cavity is not affected by the presence of ferrite. FC concluded from the measurement campaign that adding ferrites facing the beam is not feasible, that ferrites inside the wire cavities are desirable, that the forks should be terminated with a match to 50-Ohm for low frequency resonances. FC and the RLC team gave green light for the LHC wire-scanner construction. (4) Long range beam-beam and compensation simulations (UD) ---------------------------------------------------------- UD first reviewed the main concepts and constraints of the beam-beam compensation based on electro-magnetic wires. The latter include differences in the shape of the forces exerted by beam and wire, and the betatron phase advance between the long-range collision points and the compensating wire. UD then presented BBTRACK simulations for SPS and LHC, using three or four different criteria for instability: particle loss, growth in the emittance variance for a group of particle launched at the same amplitude with different betatron phases, the growth in the 4-D distance between two close-by particles, and the growth in the differences of their betatron phases. The latter two criteria are related to the Lyapunov exponent. UD also presented tune footprints for the 'stable' particles. The two Lyapunov criteria predict onset of chaos around 1.3 sigma for the SPS experiment and at about 2.6 sigma for the LHC without compensation. The onset of chaos was defined by the radius for which a circular segment contained more than 20% chaotic trajectories. At larger amplitudes, UD noticed a strong difference in the emittance growth between the LHC and the SPS, the origin of which still needs to be understood. The onset of chaos also appears to be somewhat lower than those found in full SIXTRACK simulations, where WH reported the onset of chaos to be always above 4 sigma. After the meeting, WH suggested as a possible reason the assumption of round beams, whereas the real beam shape is elliptical for about half of the long-range collision points. In case of the LHC UD performed simulations for nominal and extreme PACMAN bunches. The compensating wire raises the amplitude of chaos onset by more than a factor of 2 for the nominal bunches, whereas it assumes about the same value with or without wire for the PACMAN bunches. This suggests that a dc wire will improve the lifetime of nominal bunches without adversely affecting the PACMAN ones. Another good news is that the beneficial effect of the wire for the nominal bunches is not very sensitive to the exact wire current, but the tolerance for the latter is of order 25%. UD also presented a tune scan for nominal bunches with and without compensation, which demonstrated that the large improvement of stability by the wire is found for almost all tunes. A scan of the wire distance showed that the compensation is lost quickly for increasing distances to the beam. However, UD later realized that he may improve the result if he redoes this scan for a higher wire current. FC suggested to look at the presence and effect of an electric wire potential, which he estimated to be of order a few volts for a dc wire and 100s of Volts for a pulsed device. He also pointed out that wire vibrations might be exciting by water cooling and that these could be detrimental to the beam. FC recommended exciting the SPS wire mechanically by a loudspeaker. FZ mentioned that the RHIC wire will not be water cooled, but temperature stabilized only via heat sinks. UD's further plan is, now that a sensible stability criterion based on the phase difference is established, to perform a series of simulations for LHC, SPS, RHIC, and DAFNE. The discrepancies between SPS and LHC, as well as those between SIXTRACK and BBTRACK will be further investigated. Hardware developments and stability test measurements might be done with the help of FC. MD time at the SPS will be requested. Other possibilities are to participate in machine studies at RHIC and DAFNE. (5) SPS MD requests ------------------- GA has asked for 2006 MD requests to be submitted by 20.02.2006. FR asked for specific measurements on the LHC collimator. Suggestions for 2006 collimator measurements were: - repeating the tune shift vs. collimator gap measurement, possibly for different emittances and bunch lengths - tune shift vs. position of single collimator jaw - find beam parameters (many bunches and/or low-frequency modulation) which might allow verifying the inductive bypass effect in the real part of the impedance Other proposed MDs included the following: - measurements during e-cloud scrubbing run (e.g., tune shift along the train, stabilizing chromaticity, Schottky signals). - general SPS impedance via measurement of tune shift with current - wire compensation studies - taking 14-GeV/c data for verifying the 2004 localized impedance data - study of fast transverse instability, possibly with higher bunch intensity from the PS (6) AOB ------- FR informed the team that the publication of a new LHC electro-cloud paper by Miguel Furman is imminent. This could be taken as occasion for reviewing all aspects of the LHC electron cloud in a short note. FR recommended to redo SPS benchmarking including the rediffused electron component of Miguel Furman. EM reported a correction to the SPS BPM impedance reported in the previous week. The impedance numbers from Bruno Spataro needed to be divided by an offset of 4 mm (multiplied by a factor 250). The concrete effect of the BPM impedance will be studied in HEADTAIL simulations by GR. The presentation by EM on "Transverse RW impedance of SIS100 and LHC collimators: comparison of Zotter's and Balbekov's results" was postponed to a later meeting due to lack of time. The next RLC-LHC team meeting will be in two weeks, on 3rd March. Posted on the web: Slides by FC, UD and AG Web site: http://ab-abp-rlc.web.cern.ch/ab-abp-rlc/