Author: Gong, C.
Paper Title Page
MOPPC054 Multi-code Modelling of Momentum Collimation in the TRIUMF ARIEL Linac 253
 
  • F.W. Jones, Y.-C. Chao, C. Gong
    TRIUMF, Canada's National Laboratory for Particle and Nuclear Physics, Vancouver, Canada
 
  The electron linac component of the TRIUMF-ARIEL facility will provide CW beams of 50-75 MeV and up to 0.5 MW of beam power, with consequent requirements for low-loss operation. One factor in controlling beam quality is the reduction of the low-momentum tail arising from the rf-modulated 300 KV electron gun and initial capture elements prior to acceleration in the 10 MeV Injector linac. To study momentum collimation in the 10 MeV transfer line to the main linac, and its implications for downstream beam characteristics, a simulation model has been constructed using several tracking and optics codes, linked together by scripts and data converters. The model follows the evolution of the beam from the e-gun through the injector cryo-module and the medium energy transfer line where the proposed collimator is located. The components, methods and results of this application are described in detail.  
 
MOPPC055 A New Platform for Global Optimization 256
 
  • C. Gong, Y.-C. Chao
    TRIUMF, Canada's National Laboratory for Particle and Nuclear Physics, Vancouver, Canada
 
  Funding: Funding is received from Natural Sciences and Engineering Research Council of Canada and National Research Council of Canada for this research.
This paper describes a new platform for the multi-objective global optimization of accelerator design. While local optimization is relatively simple, global optimization for accelerator design remains a challenging task. The user often must write many lines of code to combine the output of a large variety of simulation engines, then send the results to the optimization engine. The optimization code also requires significant revision when applied to different problems. This paper presents an alternative method. The TRIUMF optimization platform, based on the genetic algorithm, is an extension of the PISA framework. It uses a flexible XML input format, in which users can easily combine multiple physics engines, such as ASTRA and PARMELA, into the same optimization problem. The TRIUMF platform is also parallel capable, designed to take advantage of computation clusters such as WestGrid. Results of the optimization platform applied to TRIUMF's 50 MeV, 0.5 MW electron linac are shown.