• M. Vretenar, C. Carli, R. Garoby, F. Gerigk, K. Hanke, A.M. Lombardi, S. Maury, C. Rossi
  CERN, Geneva

Linac4 is a 160 MeV, 40 mA H^- linear accelerator which will be the source of particles for all CERN proton accelerators from 2013. Its construction has started in 2008, as part of a program for the progressive replacement or upgrade of the LHC injectors during the next decade. Linac4 will initially inject into the PS Booster and at a later stage into a 4 GeV Superconducting Proton Linac (SPL), which could ultimately be upgraded to high duty cycle operation. For this reason accelerating structures, rf hardware and shielding of Linac4 are dimensioned for higher duty from the initial phase. Linac4 is normal-conducting, 80 m long and consists of an rf volume ion source, an RFQ, a beam chopping section and a cascade of three different types of 352 MHz accelerating structures. Its main design requirements are high reliability, high beam brightness and low beam loss. The accelerator will be housed in an underground tunnel on the CERN site, which can eventually be extended to the SPL, with equipment installed in a surface building above. The main parameters, the status of the main components, the planning, the project organisation and the civil engineering infrastructure are presented.

• M. Vretenar, Y. Cuvet, G. De Michele, F. Gerigk, M. Pasini, S. Ramberger, R. Wegner
  CERN, Geneva
• E. Kenzhbulatov, A. Kryuchkov, E. Rotov, A.G. Tribendis
  BINP SB RAS, Novosibirsk
• M.Y. Naumenko
  RFNC-VNIITF, Snezhinsk, Chelyabinsk region

The 352 MHz CCDTL will accelerate the Linac4 beam from 50 to 102 MeV. It will be the first CCDTL used in a proton linac. Three short DTL tanks, each having two drift tubes, are connected by coupling cavities and form a chain of 5 resonators operating in the stable π/2 mode. The CCDTL section is made of 7 such chains, each fed by a 1.3 MW klystron. Focusing quadrupoles are placed between tanks, easing their alignment with respect to a conventional DTL thus making the structure less sensitive to manufacturing errors. In order to validate the design and to develop the production technology, two prototypes have been constructed and successfully tested. The first prototype, built at CERN, consists of two half-cavities and one coupling cell, whereas the second, with two full cavities and one coupling cell, was built at VNIITF and BINP in Russia in the frame of an R&D contract funded by the ISTC Organisation. Both prototypes have been tested at CERN slightly beyond their nominal power level, at the design duty cycle of 10%. In this paper we present the results of high-power tests, the results of the technological developments prior to production, and the final design of the CCDTL.

• S. Ramberger, N. Alharbi, P. Bourquin, Y. Cuvet, F. Gerigk, A.M. Lombardi, E.Zh. Sargsyan, M. Vretenar
  CERN, Geneva
• A. Pisent
  INFN/LNL, Legnaro, Padova

The Drift Tube Linac (DTL) for the new linear accelerator Linac4 at CERN will accelerate H^- ion beams of up to 40mA average pulse current from 3 to 50MeV. It is designed to operate at 352.2MHz and at duty cycles of up to 10%, if required by future physics programmes. The accelerating field is 3.2MeV/m over the entire length. Permanent magnet quadrupoles are used as focusing elements. The 3 DTL cavities consist of 2, 4 and 4 segments of about 1.8m each, are equipped with 35, 41 and 29 drift tubes respectively, and are stabilized with post-couplers. Several new features have been incorporated in the basic design. The electro-magnetic design has been refined in order to reduce peak field levels in critical areas. The mechanical design aims at reducing the complexity of the mechanical structure and of the adjustment procedure. Drift tubes and holders on the tanks that are machined to tight tolerances do not require adjustment mechanisms like screws or bellows for drift tube positioning. A scaled cold model, an assembly model and a full-scale prototype of the first half tank have been constructed to validate the design principles. The results of metrological and rf tests are presented.

• C. Rossi, P. Bourquin, J.-B. Lallement, A.M. Lombardi, S.J. Mathot, M.A. Timmins, G. Vandoni, M. Vretenar
  CERN, Geneva
• S. Cazaux, O. Delferrière, M. Desmons, R.D. Duperrier, A. France, D. Leboeuf, O. Piquet
  CEA, Gif-sur-Yvette

The first stage of acceleration in Linac4, the new 160 MeV CERN H^- injector, is a 352 MHz, 3 m long Radiofrequency Quadrupole (RFQ) Accelerator. The RFQ will capture a 70 mA, 45 keV beam from the rf source and accelerate it to 3 MeV, an energy suitable for chopping and injecting the beam in a conventional Drift Tube Linac. Although the RFQ will be initially operated at low duty cycle (0.1%), its design is compatible with higher duty cycle (10%) as the front-end for a possible high-intensity upgrade of the CERN linac facility. The RFQ will be of the brazed-copper design and will be built and assembled at CERN. Beam dynamics design allows for a compact structure made of a single resonant unit. Field symmetry is ensured by fixed tuners placed along the structure. In this paper we present the rf and mechanical design, the beam dynamics and the sensitivity to fabrication and to rf errors.

• R. Garoby, O. Brunner, S. Calatroni, E. Ciapala, F. Gerigk, A.M. Lombardi, R. Losito, V. Parma, C. Rossi, J. Tuckmantel, M. Vretenar, W. Weingarten
  CERN, Geneva

As part of the upgrade of the LHC injector complex at CERN, the construction of a 4 GeV Superconducting Proton Linac (the SPL, in fact an H^- accelerator) is planned to begin in 2012. Depending upon physics requests, it should be upgradeable to 5 GeV and multi-MW beam power at a later stage. The construction of Linac4, its low energy front end, has started at the beginning of 2008. A full project proposal with a cost estimate for the low power version of the SPL aimed at improving LHC performance has to be ready for mid-2011. As a first step towards that goal, essential machine parameters like rf frequency, cooling temperature and beam current have recently been revisited and plans have been drawn for designing and testing critical components. The SPL parameters are reviewed in the context of the CERN plans for upgrading the LHC injectors, and the foreseen developments during the next years are described.

• M. Vretenar, P. Bourquin, R. De Morais Amaral, G. Favre, F. Gerigk, J.-M. Lacroix, T. Tardy, R. Wegner
  CERN, Geneva

The high-energy section of Linac4, between 100 and 160 MeV, will be made of a sequence of 12 seven-cell accelerating cavities of the Pi-Mode Structure (PIMS) type, resonating at 352 MHz. Compared to other structures used in this energy range, cavities operating in pi-mode with a low number of cells have the advantage of simplified construction and tuning, compensating for the fact that the shunt impedance is about 10% lower because of the lower frequency. Field stability in steady state and in presence of transients is assured by the low number of cells and by the relatively high coupling factor of 5%. Standardising the linac rf ystem to a single frequency is considered as an additional economical and operational advantage. The mechanical design of the PIMS will be very similar to that of the 352 MHz normal conducting 5-cell LEP accelerating cavities, which have been successfully operated at CERN for 15 years. After reviewing the basic design principles, the paper will focus on the tuning strategy, on the field stability calculations and on the mechanical design. It will also report the results of measurement on a cold model and the design of a full-scale prototype.