DEVELOPMENT OF QUARTER-WAVE CAVITIES AND FUTURE PROSPECTS FOR SUPERCONDUCTING CAVITIES

Transcription

1 EUROPEAN ORGANIZATION FOR NUCLEAR RESEARCH ORGANISATION EUROPÉENNE POUR LA RECHERCHE NUCLÉAIRE CERN - TS Department EDMS Nr: TS-Note Group reference: TS-MME 27 May 2008 DEVELOPMENT OF QUARTER-WAVE CAVITIES AND FUTURE PROSPECTS FOR SUPERCONDUCTING CAVITIES S. Calatroni Abstract A new superconducting LINAC will be constructed in the next years at CERN for the increase in energy and intensity of the REX-ISOLDE facility. The new HIE-ISOLDE LINAC will be designed and constructed in collaboration with different European Laboratories, and in its final stage will make use of 30 superconducting quarter-wave low-beta cavities of two different types. The cavities will be made with the Nb/Cu technology, on which CERN has a considerable experience, although in a different geometry, and prototypes of the lower beta ones should be developed within a couple of years. An overall view of the project will be given, followed by an illustration of the work going on and foreseen in TS/MME. A discussion of future projects involving superconducting RF at CERN (SPL, ASACUSA, EU facility for superconducting cavity test) will also be given. Presented at the 3 rd TS Workshop Archamps, France, May 27 May 29, 2008

2

3 1 INTRODUCTION RF superconducting cavities will fuel the LHC proton beams up to their final 7 TeV energy. These cavities have been produced by sputtering a thin niobium film onto a copper substrate. This technology was developed at CERN in the eighties (20 th century) and extensively applied to the former LEP collider. An intense R&D programme followed suit, based on small-scale prototype cavities, resulting in deep understanding of the technology and the underlying physics, and allowing improving performance of the coatings by a factor 3 at least. However the development was discontinued due to higher priorities within CERN. In the meantime, strong R&D was carried out worldwide for optimising the accelerating cavities for the proposed ILC collider, this time using a niobium-bulk based technology. Future CERN accelerators are projected to again make use of this technology. In particular HIE-ISOLDE will make use of low-beta quarter wave cavities, made with the Nb/Cu technology. The SPL LINAC will instead make use of low-beta elliptical cavities made with the niobium bulk technology. The ASACUSA experiment is producing a superconducting quadrupole of novel shape for trapping and cooling antiprotons. Finally, CERN is part of a bid to the EU for funding a SC cavities test facility which will make extensive use of existing facilities at CERN. All this projects will be discussed, with particular emphasis to the HIE-ISOLDE project where the TS-MME group leads the workpackage for cavity fabrication. (a) (b) Figure 1 (a) Layout of the HIE-ISOLDE accelerator in bldg 170 (in red). (b) The first modules of HIE- ISOLDE will replace some components of the existing REX-ISOLDE facility (in blue). 1

4 2 HIE-ISOLDE The HIE-ISOLDE accelerator [1] is foreseen as a booster to the present REX-ISOLDE facility in hall 170 at CERN-Meyrin. The purpose of the new facility is to increase the final energy of the radioactive isotopes produced and extracted from ISOLDE up to an energy of 10 MeV/u, with a maximum acceptable A/q of 4.5. Figure 1 illustrates the final layout of the HIE-ISOLDE facility. In the present design, two types of the quarter-wave superconducting cavities are foreseen, having geometric β of 6.26% and 10.84%. The cavities operate at a frequency of MHz, and their total number and β values have been optimised in order to provide the best compromise between total transit time factor, energy gain and A/q acceptance. Twelve and twenty cavities of each type respectively are foreseen, grouped by four or five in a cryostat which also contains a focussing solenoid. Superconducting cavities have been chosen because they allow maintaining beam quality for a wide range of beam energies, and providing the highest flexibility for operation, i.e. scaling and phasing retuning. A funded R&D program has been launched at the end of 2007 in order to prepare a full Technical Design Report covering all the issues of the HIE-ISOLDE LINAC, including cavity prototyping and testing, cryomodule design, beam dynamics and beam diagnostics [2]. This work is carried out as a collaborative effort among several European Laboratories, and is expected to be completed by end of 2009, with several deliverables including one full working superconducting cavity prototype, and full design study of the cryomodule. TS-MME is especially involved in these two workpackages, leading the first one. The choice of superconducting cavities opened the technological option for fabrication, between bulk niobium and copper coated with a niobium film. The latter has been retained, because copper can be made much more massive than niobium at significantly lower cost, with a strong gain in terms of reduction of vibrations (microphonics) which is essential for beam stability and performance. The operation of the cavities is foreseen at 4.5 K and at the target operating frequency the surface resistance should still be dominated by the residual term. The requirements in terms of peak surface electric and magnetic field are very challenging for a relatively small accelerating field (about 42 MV/m and 750 Oe for an accelerating field of 6 MV/m, high-β cavities). Although niobium films may show degradation of performances with increasing field, the power margin of 7 W allows for a surface resistance of about 100 nω, which should be feasible in this geometry. The shape of the cavities is illustrated in Figure 2 a and b. Similar cavities are in use in the ALPI accelerator at INFN-Legnaro, and the ideas developed for the HIE-ISOLDE cavities are based on their experience. A major difference will however lie in the cavity fabrication method. Whereas INFN has chosen either the option of a machined and brazed structure, or fully machined in a single copper block, we have chosen the option of basing our cavities on the well proven capabilities of CERN workshop in sheet metal working and EB-welding. Surface treatments will be performed with the well-known CERN-developed SUBU chemical polishing. First tests on a dummy machined and welded piece look promising (Ra<0.6µm), however an option for electropolishing is always open if needed. The cavity will be rinsed with ultra-pure water in clean room environment to remove dust prior to coating. The niobium coating process, on which CERN has a pioneering experience, will be based on the DC-bias sputtering technique, which has already given satisfactory performance on the INFN cavities. The parameter tuning of this coating process can however be extremely complex and time consuming, thus some testing will be performed in an existing coating facility, well before the final coating system is foreseen to be delivered to the coating lab. The coating system itself, illustrated in Figure 3, is an all-metal system, except for the Viton seals of the main flanges. This should allow reaching a pressure in the low 10-9 mbar range with a bakeout at 150 ºC. All cavity assembly work will be performed in an ISO 5 clean room. The coating jar will be closed in the clean room, and the connection to the TMP pumping station will be made through a by a vacuum valve which will allow minimising any dust contamination of the cavity surface. 2

5 According to the present planning, the first high-β cavity should be completed by end- September 2008, ready for RF validation and surface treatment testing. The coating system will also be ready by the same date, and the cathode geometry should by then have been finalised in separate testing. Samples will be coated first, in a mock-up of the high-β HIE-ISOLDE cavity, and the final objective is to perform the first test coating by the end of the year. Figure 2 (a) 3-D view of the β= cavity. (b) 3-D view of the β = cavity Figure 3 : 3-D view of the coating system, including sputtering cathodes and test cavity. The big flanges are sealed with a Viton o-ring, which could be easily replaced by a Helicoflex seal. All the other flanges are standard Conflat metal sealed flanges, to guarantee bakeability and low leak rate. Vacuum pumping will be via one of the bottom flanges. 3

6 3 SPL Figure 4 Layout of the new accelerator chain at CERN The injector chain will be the main limitation for attaining the full luminosity of the LHC collider. It is thus foreseen to replace in the mid-term (commissioning projected in 2016) the chain LINAC2 Booster PS with a full new chain LINAC4 SPL PS2. The LINAC 4 project is ready for breaking ground, and it will be followed by the SPL and PS2 in parallel. The SPL [3] LINAC will boost particles from 180 MeV up to possibly 5 GeV, and allowing for an ultimate extremely high current of 40 ma. This will be made possible by making use of a string of (at the present stage of the design) 231 low-β elliptical cavities (β=0.65 and 0.92) operating at 704 MHz, which will require extremely high performance. These performances will only be possible using bulk-niobium cavities (at the present state of the niobium-film technology), processed with advanced surface treatments such as full electropolishing, which have never been done except on β=1 1.3 GHz ILC-type cavities. R&D for the SPL is carried out within the European framework. However a strong push is being given at present to restart operation of the cavity processing plants at CERN, which have not been routinely operated for a few years. This will allow benchmarking our present capabilities, by making use of cavities lent from other Laboratories, and plan possible upgrades for being ready for cavity production with a full-fledged state-of-the art infrastructure. Support for this initiative is expected to come also from EU funds for an SC cavity facility, discussed in a following paragraph. 4 ASACUSA The ASACUSA experiment (Atomic Spectroscopy And Collisions Using Slow Antiprotons) is planning to implement a novel Paul trap based on a superconducting RF quadrupole, which will serve the dual purpose of capturing and slowing antiprotons by means of a two-mode resonating system [4]. A prototype single mode resonator is under test, for validating the principle. Its design has been also based on the expertise of TS-MME in superconducting RF and related surface treatments. Design work for the full system is well undertaken, and fabrication of the cryostat by the CERN workshop should be completed by end In parallel, development work for the full resonator is advancing, and an informal study group with TS-MME involvement has allowed the project to iron out a few items. This project will also profit in future from a better integration within the re-emerging CERN SC cavity community, which will be centred on the SC cavity main test facility. 4

7 5 SC CAVITIES FACILITY AT CERN The focussing of all DESY resources on the challenging XFEL project has left the European SC cavity community without a testing facility of easy access, and it has thus been decided by the CERN Council Strategy Group to study the feasibility of a central European facility located at CERN [5]. The study of this facility has undergone considerable discussion during the past two years, and the agreement for a multi-site facility resulted in a bid to the EU at the level of 5.8 MEUR, which is expected to be approved soon for funding starting in CERN will benefit from more than half of the amount, and a good part of it is earmarked for improving the surface treatment facilities. This will be of course of considerable help for all the projects listed above, and will be a sure catalyser for all the activities and for people s involvement. 6 CONCLUSIONS RF superconductivity has been instrumental in both the largest recent CERN projects, namely the LEP and LHC colliders. Although the activity has been kept dormant in the last few years, because of other priorities related to the LHC, several new projects using SC cavities are recently taking shape. TS-MME has been involved in all these projects from the earliest stages, and could provide valuable knowledge in the design phase. Several projects are now nearing the hardware manufacturing stage, and again highly specialised skills have to be revived in order to bring all of them to successful completion. ACKNOWLEDGEMENTS The author wishes to thank M. Pasini and M. Lindroos (AB Dept.) and all TS-MME people involved in the HIE-ISOLDE project, namely L. Ferreira, A. Mongelluzzo, F. Thierry and Ph. Trilhe. R. Garoby, F. Gerigk, R. Losito (SPL), M. Hori, W. Pirkl (ASACUSA) and W. Weingarten (SC-RF facility) are also gratefully acknowledged for sharing key information. REFERENCES [1] HIE-ISOLDE: the technical options, M. Lindroos and T. Nilsson Eds., CERN [2] [3] [4] The ASACUSA Collaboration, CERN/SPSC [5] 5