Abstract:Muon scattering imaging technology can be used to detect nuclear material and is of considerable significance in nuclear safety. However, it is difficult to distinguish special nuclear materials from high-Z objects effectively by using the existing muon scattering imaging technologies. Muon-induced neutrons emitted from special nuclear materials can help to identify the existence of special nuclear materials. However, this method has long imaging time and low imaging quality. Multimodal imaging of muon uses both the information about scattering muons penetrating the material and the information about muons stopped by material and generating secondary induced neutrons, which can overcome the shortcomings of single imaging method effectively. The detection model is set up based on Geant4. The simulation programs of muon imaging in coincidence with muon induced neutrons, scattering imaging of muon, and multimodal imaging of muon are developed by using Cosmic-ray Shower Library as particle source, and the imaging algorithms are implemented respectively on the basis of the simulated data. Two imaging models are designed for muon scattering imaging. The first one is a single 235U cube, and the second one is composed of four cubes, namely 235U cube, 239Pu cube, lead cube and aluminum cube. This simulation has completed muon scattering imaging of single cube and four cubes. In the part of muon imaging in coincidence with muon induced neutrons, the neutronic gain of the HEU (90% 235U) plate, LEU (20% 235U) plate, and DU (0.2% 235U) plate, as well as the relationship between the neutronic gain of these three uranium plates and the energy and charged properties of the muon are obtained by simulation, and then two imaging models are set up. The first one is composed of four cubes, namely 235U cube, 239Pu cube, lead cube, and aluminum cube, and the other is comprised of multilayer nuclear components. The 2D and 3D reconstruction results of multi-objects and multilayer nuclear components are obtained through muon imaging in coincidence with muon induced neutrons. Then the multimodal imaging of muon for three cubes is realized in the presence or absence of iron shielding shell. The imaging capabilities are compared with the muon scattering imaging capacities and muon imaging capacities in coincidence with muon induced neutrons. Simulation studies indicate that multimodal imaging of muon based on scattering and secondary induced neutrons can effectively combine the advantages of every single imaging method. The multimodal imaging of muon can take advantage of available information more efficiently, which is helpful in improving the imaging quality. Multimodal imaging of muon not only has the advantages of short imaging time and high imaging quality, but also can distinguish special nuclear material from other high-Z materials clearly, which is vital for detecting special nuclear materials. Keywords:cosmic ray muon/ muon scattering imaging/ muon induced neutrons/ multimodal imaging of muon
模拟结果如图7所示, 天然宇宙线缪子能谱下HEU板统计得到的中子数目较LEU板和DU板均有大幅增加, 这是由于235U有更大的裂变截面, 使缪子诱发裂变产生了更多的次级中子. 在特定能量缪子打靶时, HEU板仍然较LEU板、DU板产生更多的次级中子; 在三种235U丰度的铀板中, 低能负缪子诱发裂变产生的中子数相比其他设定参数下产生的中子数具有数个量级的优势. 更强的次级信号增加了通过诱发中子标记与其符合的缪子的可能性, 为在高原子序数材料中区分SNM提供了可靠途径. 图 7 中子增益结果 Figure7. Result of neutronic gain.
24.2.模型设置 -->
4.2.模型设置
设置了两种成像模型, 模型一与图3(b)所示模型相同, 仍通过CRY发射500万个缪子. 模型二如图8所示, 展示了根据国外典型核武器理论模型设计的简单核部件结构[26-28], 由内到外依次为空腔、239Pu层、Be层、238U层、高能炸药层(黑索金C3H6N6O6)、Al外壳. 图8(a)为模型具体结构, 图8(b)为Geant4建模结果(为展示部件具体组成, 以半球显示). 核部件结构相对精细, 模拟使用CRY发射1000万个缪子. 图 8 诱发中子符合的缪子成像模型 (a) 核部件具体结构图; (b) 核部件建模示意图 Figure8. Imaging model of muon imaging in coincidence with muon induced neutrons: (a) Detailed structure diagram of nuclear components; (b) nuclear components model.
24.3.成像结果 -->
4.3.成像结果
根据成像原理, 分别得到了两种模型在Z = 0平面的二维成像结果以及在不同Z值处断层扫描得到的三维成像结果, 如图9和图10所示. 二维成像平面大小为1000 × 1000, 像素大小为10 × 10, 第j个像素网格内ρj越大, 网格颜色越深, 三维成像由序列连续的多张二维图像融合得到. 图9(a)和图9(b)为模型一成像结果, 据二维成像图可知, 利用与诱发中子符合的缪子能够明显区分SNM与其他材料, 成像显示的四个物块所在位置以及239Pu块、235U块大小较为准确, 且239Pu块较235U块成像更加均匀, 物块边界也更加清晰; 三维成像结果较为立体地重建了239Pu块和235U块. 图10(a)和图10(b)为模型二的成像结果, 据二维成像图可知, 核部件整体形状成像较好, 外部的高能炸药和Al壳与内部的反射层分辨清晰, 最内层空腔与裂变材料239Pu层分辨清晰, 239Pu层与反射层边界较为模糊; 三维成像图由Z ≥ 0的二维断层重建图像得到, 成像结果显示了239Pu层与238U层之间的区域. 图 9 四物块模型的诱发中子符合的缪子成像图 (a) 四物块模型二维成像结果; (b) 四物块模型三维成像结果 Figure9. Muon imaging in coincidence with muon induced neutrons of four cubes model: (a) 2D imaging results of four cubes model; (b) 3D imaging results of four cubes model.
图 10 核部件模型的诱发中子符合的缪子成像图 (a) 核部件模型二维成像结果; (b) 核部件模型三维成像结果 Figure10. Muon imaging in coincidence with muon induced neutrons of nuclear components: (a) 2D imaging results of nuclear components; (b) 3D imaging results of nuclear components.
-->
5.1.模型设置
多模态成像设置了两种模型, 模型一如图11(a)所示, 沿X轴由上至下等间距放置了边长均为10 cm的235U块、铅块、铝块. 进一步设置了如图11(b)所示的屏蔽模型二, 以检验成像方法在有5 cm厚的铁屏蔽外壳存在的情况下, 探测SNM的能力. 模型一、模型二通过链接CRY分别发射200万个缪子和500万个缪子. 图 11 缪子多模态成像模型 (a) 三物块成像模型; (b) 铁外壳屏蔽的三物块成像模型 Figure11. Imaging model of multimodal imaging of muon: (a) Three cubes model; (b) three cubes model with iron shielding shell.
图 13 铁屏蔽下三物块成像结果 (a) 缪子散射成像; (b) 诱发中子符合的缪子成像; (c) 缪子多模态成像 Figure13. Imaging results of three cubes model with iron shielding shell: (a) Muon scattering imaging; (b) muon imaging in coincidence with muon induced neutrons; (c) multimodal imaging of muon