Multiple-charge-state Beam Steering in High-intensity Heavy-ion Linacs*

An algorithm suitable for correction to steering of multiple-charge-state beams in heavy-ion linacs operating at high currents has been developed [1]. It follows a fourdimensional minimization procedure that includes coupling of the transverse beam motions. A major requirement is that it obeys the restricted lattice design imposed by the acceleration of multiple-charge-state heavy-ion beams [2]. We study the algorithm efficiency in controlling the beam effective emittance growth in the presence of random misalignments of cavities and focusing elements. Limits on misalignments are determined by quantifying beam losses and effective steering requirements are selected by examining several correcting schemes within the real state constraints. The algorithm will be used to perform statistically significant simulations to study beam losses under realistic steering. MULTI-Q-STATE BEAM DYNAMICS For heavy-ion linacs an efficient way to achieve the high current intensities required by a high-power mediumenergy machine is to accelerate simultaneously several charge states of stripped ions. Simultaneous acceleration of multiple-charge states overcomes the limitations of presently available ion sources and allows the use of a larger portion of the stripped beam. In addition, by using multiple strippers the linac length necessary to achieve the design energy can be reduced. Large transverse and longitudinal acceptances and minimization of drift spaces between cavities are effective means to preserve the beam quality. Other sources of emittance dilution, such as misalignments of transverse focusing elements and resonators can lead to beam losses and require correction. A steering algorithm designed for multiple-charge-state beams and obeying the lattice limitations and machine complexities of a high-intensity heavy-ion linac has been developed and preliminary results were published in [1]. Here, we present results obtained with a fully-developed version of the algorithm code implementation, applied to the linac driver of the proposed Rare Isotope Accelerator (RIA) project [2]. The RIA linac driver has a high degree of complexity due to the large number of components and the requirement that it accelerates beams of any ions, including uranium, to energies up to 400 MeV/nucleon and 400 kW of beam power. ____________________________________________ * Work supported by the U.S. Department of Energy, Office of Nuclear Physics, under Contract No. W-31-109-ENG-38. [email protected] MINIMIZATION An effective steering algorithm for multi-q ion beams should control emittance growth and reduce trajectory excursions to avoid beam losses. Most importantly, it should be tailored to the restricted choices of steering and diagnostics configurations, so as to be implement-able in a real machine. In the lower-energy sections of a SC heavy-ion linac designed to accelerate multi-q beams correctors need to be closely spaced, with more than one corrector placed in the same cryostat. Beam position monitors (BPMs) are placed between cryostats. Correction methods that zero out the beam position at a monitor by varying an upstream corrector tend to fail, because in general for such accelerators, the linear transport matrices between correctors and monitors form singular systems that require appropriate mathematical tools. A many-correctors-to-one-monitor system is best solved by least-square minimization. Our algorithm is based on the determination of the beam response functions to known induced excitations to the beam trajectory and on the minimization of a goal function that depends on those transfer functions. We assume that the beam centroid can be mapped by functions relating the initial phase-space coordinates at a point s0 to its coordinates at a point s along the accelerator. These transfer functions describe the lattice responses at s to the beam conditions at s0. Given N misaligned elements, we need 2N+4 measurements of the beam position and angle at the BPMs to determine the misalignments and initial conditions exactly. Having more unknowns than equations, we find the corrector strengths by minimizing the function: