会议论文详细信息
International Workshop on Neutron Optics and Detectors
Optimization of multi-channel neutron focusing guides for extreme sample environments
Di Julio, D.D.^1 ; Lelièvre-Berna, E.^1 ; Courtois, P.^2 ; Andersen, K.H.^1 ; Bentley, P.M.^1,3
European Spallation Source ESS AB, PO Box 176, SE-221 00 Lund, Sweden^1
Institut Laue-Langevin, 6 rue Jules Horowitz, 38042 Grenoble Cedex 9, France^2
Department of Physics and Astronomy, Uppsala University, Uppsala, Sweden^3
关键词: Artificial bee colonies;    Central channels;    Differential Evolution;    Monte-Carlo ray tracing;    Neutron focusing;    Optimization algorithms;    Population-based algorithm;    Sample environment;   
Others  :  https://iopscience.iop.org/article/10.1088/1742-6596/528/1/012006/pdf
DOI  :  10.1088/1742-6596/528/1/012006
来源: IOP
PDF
【 摘 要 】

In this work, we present and discuss simulation results for the design of multichannel neutron focusing guides for extreme sample environments. A single focusing guide consists of any number of supermirror-coated curved outer channels surrounding a central channel. Furthermore, a guide is separated into two sections in order to allow for extension into a sample environment. The performance of a guide is evaluated through a Monte-Carlo ray tracing simulation which is further coupled to an optimization algorithm in order to find the best possible guide for a given situation. A number of population-based algorithms have been investigated for this purpose. These include particle-swarm optimization, artificial bee colony, and differential evolution. The performance of each algorithm and preliminary results of the design of a multi-channel neutron focusing guide using these methods are described. We found that a three-channel focusing guide offered the best performance, with a gain factor of 2.4 compared to no focusing guide, for the design scenario investigated in this work.

【 预 览 】
附件列表
Files Size Format View
Optimization of multi-channel neutron focusing guides for extreme sample environments 630KB PDF download
  文献评价指标  
  下载次数:15次 浏览次数:25次