Nuclear Techniques ›› 2018, Vol. 41 ›› Issue (3): 30201-030201.doi: 10.11889/j.0253-3219.2018.hjs.41.030201

• LOW ENERGY ACCELERATOR, RAY AND APPLICATIONS • Previous Articles     Next Articles

The assessment for the effects of multiple combined variance reduction techniques in shielding computation on an electron irradiation accelerator

CUI Tiantian1,2, ZHANG Shuyuan1, SUN Liang1, LU Jieping1   

  1. 1. School of Radiological Medicine and Protection, Medical College of Soochow University, Suzhou 215123, China;
    2. CGN Dasheng Electron Accelerator Technology Company, Suzhou 215214, China
  • Received:2017-05-18 Revised:2017-09-29 Online:2018-03-10 Published:2018-03-14
  • Supported by:
    Supported by Project of International Thermonuclear Experimental Reactor (No.2014GB112006), National Natural Science Foundation of China (No.11575124)

Abstract: [Background] Variance reduction is a common method to resolve the problems in deep penetration application such as radiation shielding, medical physics and well-logging. Currently, the study of the variance reduction techniques is based on simple theoretical geometric model, lacking the discussion on complex practical applications. [Purpose] This paper aims to find an optimized combined variance reduction technique in deep penetration application on the practical complex model. [Methods] With an electron irradiation accelerator as the model, this paper studied the effects of combined variance reduction techniques on MCNP5 (Monte Carlo N Particle Transport Code) platform, including geometry splitting, biasing the source, exponential transform, step-step transform, DXTRAN and weight window. [Result] The results showed that the irradiation products had a great influence on the angular distribution of spectrum of the new source, and the geometric model should not be overly simplified. Compared with other variance-reducing techniques, the step-step transform and biasing the source efficiency reduced the relative error and could get a better calculation efficiency combined with geometric splitting, exponential transform and DXTRAN. Weight window had better computation efficiency and more objectivity. [Conclusion] The step-step transform and biasing the source, combined with other variance reduction techniques can resolve deep penetration problem in complex practical application based on weight window.

Key words: Deep penetration, Variance reduction, Monte Carlo, Electron irradiation accelerator

CLC Number: 

  • TL99