RFQ Designs and Beam-Loss Distributions for IFMIF

Jameson, Robert A

Oak Ridge National Laboratory

关键词: Radioactivity;    Radiation Protection;    Minimization;    43 Particle Accelerators;    Shielding;   

DOI  :  10.2172/931554
RP-ID  :  ORNL/TM-2007/001
RP-ID  :  DE-AC05-00OR22725
RP-ID  :  931554

美国|英语

来源: UNT Digital Library

【 摘 要 】

The IFMIF 125 mA cw 40 MeV accelerators will set an intensity record. Minimization of particle loss along the accelerator is a top-level requirement and requires sophisticated design intimately relating the accelerated beam and the accelerator structure. Such design technique, based on the space-charge physics of linear accelerators (linacs), is used in this report in the development of conceptual designs for the Radio-Frequency-Quadrupole (RFQ) section of the IFMIF accelerators. Design comparisons are given for the IFMIF CDR Equipartitioned RFQ, a CDR Alternative RFQ, and new IFMIF Post-CDR Equipartitioned RFQ designs. Design strategies are illustrated for combining several desirable characteristics, prioritized as minimum beam loss at energies above ~ 1 MeV, low rf power, low peak field, short length, high percentage of accelerated particles. The CDR design has ~0.073% losses above 1 MeV, requires ~1.1 MW rf structure power, has KP factor 1.7,is 12.3 m long, and accelerates ~89.6% of the input beam. A new Post-CDR design has ~0.077% losses above 1 MeV, requires ~1.1 MW rf structure power, has KP factor 1.7 and ~8 m length, and accelerates ~97% of the input beam. A complete background for the designs is given, and comparisons are made. Beam-loss distributions are used as input for nuclear physics simulations of radioactivity effects in the IFMIF accelerator hall, to give information for shielding, radiation safety and maintenance design. Beam-loss distributions resulting from a ~1M particle input distribution representative of the IFMIF ECR ion source are presented. The simulations reported were performed with a consistent family of codes. Relevant comparison with other codes has not been possible as their source code is not available. Certain differences have been noted but are not consistent over a broad range of designs and parameter range. The exact transmission found by any of these codes should be treated as indicative, as each has various sensitivities in its internal methods. Continued work to compare results between different codes more broadly and deeply than heretofore is highly recommended - this requires comparison at source code level and devising of appropriate tests. It is strongly recommended that the project obtain source code for all important simulation work.