MD 2485: Active halo control using narrowband and colored noise excitations

Transcription

1 CERN-ACC-NOTE February 2018 MD 2485: Active halo control using narrowband and colored noise excitations H.Garcia-Morales, Royal Holloway University of London, United Kingdom R. Bruce, G. Kotzian, E. Maclean, S. Redaelli, D. Valuch,J. Wagner, CERN, Geneva, Switzerland Keywords: LHC, collimation, ADT. coloured noise, active halo control, halo, detuning with amplitude Summary This MD note summarizes the actions carried out during the MD 2485 on Active halo control using narrowband and colored noise excitations. The goal of the MD was to repeat some promising cases already tested in the past and introduce a new excitation type based on applying a colored noise. Although we were able to repeat some cases using a narrowband excitation, due to a problem with the waveform generator, the colored noise excitation could not be accomplished as expected. In any case, we provide some results that may be useful for future MDs. Contents 1 Introduction and motivation 2 2 MD procedure and overview 2 3 Results Detuning with amplitude Collimator scrapings Core Tail Summary and conclusions 6 5 Acknowledgements 7 1

2 1 Introduction and motivation This MD investigates possible mitigations for spurious beam dumps due to overpopulated beam tails being scraped instantaneously on the collimators in case of orbit jitter. The investigated techniques aim at providing an active halo control to deplete the tails in a safe and controlled manner. The results could, depending on the operational bottlenecks encountered, be very beneficial both for Run III and HL-LHC (including machine protection aspects during fast crab-cavity failures). This MD aims at testing one of the techniques of halo depletion that uses the present hardware based on ADT narrow-band excitation with a bunch-to-bunch control. The main goal of this MD is to perform a beam tail scraping without affecting the core distribution and to evaluate the halo population. A dependence of tune on particle amplitude is required for the method to work. The aim is to find appropriate modulation frequency to drive resonances on halo particles only while leaving the beam core unaffected. This MD builds upon the investigations of MD312 and MD1388 [1, 2] on the active halo control which were performed at injection during MD block 2 and and in 2016 in block 3 and 4. The first attempts testing this novel technique showed that a bunch-by-bunch action with the LHC transverse damper (ADT) to excite at certain tunes is possible. The method to deplete beam tails in a controlled manner needs further tests. The previous results by using the ADT in a wide range of tunes but the effect on the beam was not really clear. A more selective frequency usage could lead to improved results. The main priority of the MD is to repeat the one case of halo depletion from the first MD, which showed promising results, but with larger excitation period. In this case, a fixed ADT frequency corresponding to a fractional horizontal tune of was used. As a novelty this MD will test for the first time the use of coloured noise for exciting a whole tune region at a time simultaneously as proposed during the HiLumi meeting in We expect to achieve much higher depletion rates with this operation mode of the ADT system. 2 MD procedure and overview The MD was scheduled to last for 7 hours the 30th of November in parallel with MD2720 on BLM spike patter recognition. Beam 1 was used in this MD. There was a initial delay of 4 hours due to an access to replace one of the octupole power supplies. The time lost was shared with the following MD. Before starting with the ADT excitations, we measured the amplitude detuning in order to have a precise correspondence between excitation frequency and excited amplitude. The measurements was performed following the AC dipole kick method. The beam is excited transversally at different amplitudes. After the detuning with amplitude measurement, we started the injections to study the 2

3

4 σ 2σ 3σ 4σ 5σ Q x AC-dipole detuning Free oscillations J x [µm] Figure 2: Horizontal detuning with amplitude fir free oscillations and for AC-dipole driven excitation. 3 Results 3.1 Detuning with amplitude In Fig. 2 the detuning with amplitude is shown as a function of the horizontal action J x. The dots represent the measurement carried out during the first hour of the MD and the red line is the fit if the AC-dipole driven detuning with amplitude. The slope of the free oscillation detuning with amplitude is a factor 2 smaller than the measured one with the AC-dipole. We took this result as a reference for selecting the ADT frequencies to excite precise amplitudes of the beam distribution for both, fixed frequency excitation and colored-noise excitation. 3.2 Collimator scrapings To extract the detailed transverse profile of the bunches, we performed collimator scrapings using one jaw of the horizontal primary collimator (TCP.C6L7.B1). The scraping was performed in 50 µm every 3 seconds until the full beam was scraped. The recorded BLM signal during the scraping is used to reconstruct the beam profile correlating it with the collimator position. In Fig. 3 the BLM signal as a function of the TCP position is shown for the different scrapings. In Fig. 4 the reconstructed profiles are shown in log scale. We split the study in two different parts. One is devoted to the study of the effects of the excitation on the core and the other one focuses on the effect on the tails Core In all the injections where we studied the core, before starting the excitation, the tail of the beam was scraped until 3σ. In such a way we ensure that the tail is not populated 4

5

6 beforehand and that the repopulation after that is mainly due to the external excitation. In Fig. 4 the core reference scraping is represented by the orange dots while the reconstructed core after an excitation using a fixed ADT frequency of is represented by the red dots. In the later case, one can see how there is a clear tail repopulation due to the excitation and an emittance blow up as it was observed online using the BSRT data. From this result we can conclude that the core is clearly affected by the excitation Tail In order to populate the beam tails we use a white noise excitation given by the ADT to blow up the individual bunches. The emittance was artificially increased by a factor 4 (until a value around 6 µm). This allows a more precise study of the effects of the different excitations on the tail distribution. In Fig. 4, three different tail profiles are shown. The blue dots represent the tail profile when no excitation is applied. This is used as a reference for following tail profiles after excitation. The green dots represent the reconstructed profile after a fix frequency excitation of which corresponds to an amplitude of about 5.5σ. We observe a small reduction on the tail population with respect to the reference case. A second profile is shown after testing different colored-noise excitations between 4σ and 6σ. Within this range, the excitation amplitude is linearly increasing. The unique injection devoted to this type of excitation we could perform was mainly for set it up and to find the most suitable set of parameters for the following injections. Unfortunately, after this set up injection there was a problem with the waveform generator and we could not apply this type of excitation anymore during this MD. Nevertheless, we reconstructed the profile of this beam using the collimator scraping. In this case there is a clear reduction of the tail population. It is difficult to conclude anything from this result since several excitations were applied to find the proper set up and the bunch population was largely reduced due to excitations with too high amplitude. Therefore, we cannot extract any conclusions of the effects of colored noise excitations yet. 4 Summary and conclusions In this MD we tested again a narrowband excitation using the ADT for halo depletion. During the first hour of the MD, the detuning with amplitude was measured using the AC dipole. Then, we successfully applied fixed frequency ADT excitations which were already good candidates from previous MDs. Finally, a colored-noise excitation was applied for the very first time. Unfortunately, we had a problem with the waveform generator which creates the colored noise excitation and the test could not be completed. Nevertheless, we obtained interesting information for future tests using this new type of excitation. 6

7 5 Acknowledgements WewouldliketothanktheOPteamoperatingthemachineduringtheMDandtheLHC-MD coordinators for its flexibility extending the MD time. References [1] J.Wagner et al. Active halo control through narrow-band excitation with the ADT, CERN, Geneva, Switzerland, Rep. CERN-ACC-Note , Apr [2] J.Wagner et al. Active halo control through narrow-band excitation with the ADT at injection, CERN, Geneva, Switzerland, Rep.CERN-ACC-NOTE [3] T. Mastoridis Crab cavity RF noise mitigation and transverse tail cleaning, HiLumi meeting May