Efficient Coherent Mode Decomposition for Physical Optics Simulations of Beamlines and Experiments

Author:

Li Ruizi,Chubar Oleg

Abstract

AbstractWe present simulations of the partially coherent undulator radiation (UR) propagation through several hard X-ray beamlines at NSLS-II, performed with SRW code, adopting two different memory- and CPU-efficient strategies / algorithms for the coherent mode decomposition (CMD). One of these algorithms employs the subtraction of quadratic phase terms from 4D cross spectral density (CSD) at a non-zero radius of the radiation wavefront curvature, whereas the other one applies the CMD at / near a radiation beam waist, where the radiation wavefront is typically “flat”. We show that, when performing the CMD on the 4D CSD of the partially coherent radiation wavefront, the required numbers of grid points of radiation meshes are largely reduced in both cases, compared to standard approaches. This dramatically reduces memory requirements (from TBs typically to several GBs) and improves the time-to-solution of the decomposed modes (from many hours at a parallel execution to several minutes at a sequential execution). Thanks to the implemented algorithms, the propagation simulations of tens or hundreds of thousands coherent wavefronts generated by independent electrons for an undulator-based X-ray beamline at a low- or intermediate- emittance synchrotron light source is reduced to the simulation of over fewer than one thousand major coherent modes obtained from the CMD, which fully recovers the structure of the partially coherent UR before and after the propagation through a beamline. The efficiency and accuracy of the methods is further demonstrated with simulations for two hard X-ray NSLS-II beamlines: HXN and CHX.

Publisher

IOP Publishing

Subject

Computer Science Applications,History,Education

Cited by 1 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3