3-D gamma radiation field reconstruction method using limited measurements for multiple radioactive sources. (15th September 2022)

• Record Type: Journal Article
• Authors: Zhu, Shangzhen; Zhuang, Shuhan; Fang, Sheng; Dong, Xinwen; Li, Wenqian; Zhang, Lei; Liu, Xinkai; Li, Ren; Liu, Xuegang; Cao, Jianzhu
• Is Part Of: Annals of nuclear energy. Volume 175(2022)
• Publication Date: 2022-09-15
• Subjects: Gamma radiation field -- Three-dimensional reconstruction -- Simulations and validations with multiple sources; Nuclear energy -- Periodicals; Nuclear engineering -- Periodicals; 621.4805
• DOI: 10.1016/j.anucene.2022.109247
• Languages: English
• ISSNs: 0306-4549

Abstract

Highlights: Capable of reconstructing the 3-D radiation field of different numbers of radiation sources. Works with both randomly and regularly sampled data. Verified with multiple Monte Carlo simulations and validated against a 1-D experiment. Takes no more than 3.125% of simulation data for 3-D field reconstruction. Mean relative errors can reach 2.18% for simulation and 7.87% for experiment. Abstract: The reconstruction of 3-D gamma radiation field is important for radiation protection in the presence of unknown radiation sources and limited measurements, such as the decontamination and decommissioning of nuclear facilities. While there may be multiple radioactive sources in these applications, the current 3-D reconstruction methods do not take the multiple radiation sources into consideration. This paper develops a method that has the potential to reconstruct the 3-D gamma radiation field of multiple radioactive sources with limited measurements of the field. This method solves a 3-D fourth-order TV-H -1 equation to promote the continuity of the reconstructed 3-D radiation field, which simultaneously fills the missing dose rates and recovers local details near the radioactive sources. The proposed method was verified against three Monte Carlo simulations with different numbers of sources and validated against a 1-D experiment. The method was tested with different numbers of combinations of the input data, including both randomly and regularly distributed data sets. The results reveal that the method accurately reconstructs the 3-D radiation field and achieves a relative error level as low as 2.18%, using only 3.125% of the data for the simulation cases. The 1-D experiment shows a relative error range between 5.57% and 65.28%. These results preliminary demonstrate the potential of the proposed method for 3-D radiation field reconstruction in practice.