Simulation study of a high-intensity linear accelerator operated at longitudinal phase advances above 90°

Abstract

High intensity beams may suffer from serious beam quality degradation if the focusing scheme allows for occurrence of resonances or instabilities. For transverse focusing a commonly accepted and respected lattice design rule is to choose the phase advance per structure period below the 90\ifmmode^\circ\else\textdegree\fi{} resonant stop band, which was implicitly applied to the longitudinal phase advance the same way. A recent study pointed out that for lattice structures with more than one rf gap per period the 90\ifmmode^\circ\else\textdegree\fi{} restriction needs not to be applied the same way to the longitudinal focusing as to the transverse one [I. Hofmann and Oliver Boine-Frankenheim, Phys. Rev. Lett. 118, 114803 (2017)], thus offering more design flexibility. The present paper is motivated by an interest to accelerate intense proton beams above longitudinal 90\ifmmode^\circ\else\textdegree\fi{} in the new poststripper drift tube linac of the GSI universal linear accelerator by using the rf power supplies optimized for the much stiffer heavy ions. We confirm that a strictly periodic prolongation of the first cavity could allow acceleration above longitudinal 90\ifmmode^\circ\else\textdegree\fi{} with only minor beam quality degradation. However, the combination of a high longitudinal phase advance with intercavity sections breaking the periodicity shows that matching challenges---rather than resonances---determine emittance growth. In the present case it is found tolerable for up to three such sections. This confirms the validity of the principle and at the same time its limits under practical conditions.