Chemical composition analysis for X-ray transport container scans. A. Zelenaya 1 , M. Zelenyi 1 , 2 , A.A.Turinge 1 , V.G. Nedorezov 1 1 Institute for Nuclear Research RAS 2 Moscow Institute of Physics and Technology (SU) October 10, 2018
Introduction ◮ It is important for national security to control the movement of dangerous or strategically cargo. ◮ This control can be provided by scanning transport containers by gamma rays produced by bremsstrahlung. ◮ In this report we consider: ◮ Methodology, existing solution and our proposing method; ◮ GEANT4 simulation of gamma rays scanning; ◮ Measurement resolution of gamma rays detector. 2 / 13
Methodology: Gamma ray attenuation T - transmittance S ( E 0 , E ) - response E 0 function � S ( E 0 , E ) exp ( − µ ( E , Z ) × t ) dE ) 0 µ ( E , Z ) - attenuation T ( E 0 , t , Z ) = E 0 t - optical thickness � S ( E 0 , E ) dE E 0 - up-limit energy of 0 bremsstrahlung E - energy of gamma ray Attenuation curve Z - charge of nuclei Z = 5 Bremsstrahlung spectrum Z = 13 by 10 MeV electron Z = 26 10 0 1 10 , cm 2 Z = 82 gr Probability density 10 1 10 2 10 0 10 1 10 3 Energy, MeV 0.0 2.5 5.0 7.5 10.0 Energy, MeV 3 / 13
Methodology: Existing solution Dual energy method ◮ We can’t defined Z if optical thickness t is unknown. ◮ But we can use two electron beams with different energy which give gamma rays with up-limit energy E ( 1 ) and E ( 2 ) 0 . 0 ◮ Then we can get Z as a result of minimizing this function: F ( z ) = | t ( E ( 1 ) 0 , z ) − t ( E ( 2 ) 0 , z ) | → min t ( E ( 1 ) 0 , z ) ◮ This technique allows to determine scan object as a one from four possible groups: Z eff ∼ 5, Z eff ∼ 13, Z eff ∼ 26, Z eff ∼ 82. 4 / 13
Methodology: Disadvantages and proposing method Disadvantages of dual energy method ◮ It is too difficult to irradiate the target with beams with different energy. ◮ Low efficiency for target which contains elements with strongly different charges. Our method ◮ Use only one electron beam with energy E = 10 MeV . ◮ Measure not only the space distribution, but also the energy of gamma rays. 5 / 13
Simulation: Preliminary estimates Steel container Size: 2x2 m Detector Thinknes: 2 mm Pixel size: 1x1 cm W γ Dangers γ e - 10 MeV γ 6 / 13
Simulation: Preliminary estimates Uranium cube (6cm) in a lead Very dangerous item sphere (thickness – 1 cm) 100 75 50 25 y, cm 0 25 50 75 100 100 50 0 50 100 x, cm 7 / 13
Simulation: Preliminary estimates Comparison the energy spectrum for aluminium and uranium orb (radius – 1 cm). Search of the explosive Logarithm ratio of intensity 2 1 0 0.0 2.5 5.0 7.5 Energy, MeV 8 / 13
Simulation: Preliminary estimates N – number of detector cells The energy deposit in detector N [ E dep < 3 MeV ] cells for several materials N [ E dep > 3 MeV ] 9 / 13
Simulation: Thickness reconstruction for the 1-D case Simple model Detector Attenuation of gamma ray Z ~ 13 Z~ 26 Z ~ 82 flux is defined as: Gamma ray N ( E ) source � Σ mean N 0 ( E ) = exp ( − ( E ) x i ) i i Reconstruction algorithm where x i — thickness of the i -layer, Σ mean — mean i ◮ Full thickness is known. cross-section for group of ◮ Find thickness using least materials, N , N 0 — the squares: number of gamma. ( ln N ( E ) ◮ Disregard secondary ( E ) x i )) 2 → min � � Σ mean N 0 ( E )+ i scattering. i E ◮ Disregard the annihilation line. 10 / 13
Thickness reconstruction: Example 150 We consider the object Original Al which has 3 layers of Fe aluminium, iron and lead. 100 N 0 ( E ) Pb ln N ( E ) The figure shows a contribution of every 50 reconstructed material in summary attenuation. 0 The table contains the 0 2 4 6 8 10 results of reconstruction. Energy, MeV Material Real, cm Reconstructed, cm Al 20 19.6 Fe 40 41.6 Pb 30 28.7 11 / 13
Simulation: Thickness reconstruction for the 1-D case Also we conducted several numeric experiments for Distribution of reconstruction errors different thickness. for various numerical experiments 30 % ◮ Aluminium, Iron and 25 % Lead are used. Relative error 20 % ◮ Several sets of 15 % thickness with full thickness from 30 cm 10 % to 180 cm. 5 % ◮ The energy grid 0 % Z ~ 13 Z ~ 26 Z ~ 82 spacing emulates the detector with 10 % resolution. 12 / 13
Simulation: Perspectives ◮ If we use energy-space distribution we can develop the algorithm for the 3D-tomography Detector Z ~ 13 Z~ 26 Z ~ 82 Gamma ray source 13 / 13
Experiment: Energy resolution of the detector The experiment was conducted by Dr. Guber and Dr. Ivashkin. 14 / 13
Experiment: Energy resolution of the detector 22 Na — 0 . 511 MeV 22 Na — 1 . 275 MeV 137 Cs — 0 . 662 MeV The sum of signals from 2 photodiode Noise threshold: 100 KeV Energy, MeV Sigma/Mean 0.511 14.7% 0.662 19% 1.275 13% 15 / 13
Conclusion Results 1. The measurement of the gamma ray spectrum allows to identify cargo belonging to the group of materials with certain Z eff . 2. Also it allows to define the thickness of layers from different elements with the accuracy about 25%. 3. The energy resolution of the detector based on a BGO scintillator was studied. For the photodetector with full array of pixels the energy resolution is expected about 10%. Plans and perspectives With financial support can be developed: 1. The program which checks cargo of a transport container for compliance cargo manifest 2. The algorithm for the 3D gamma-tomography. 16 / 13
Thank you for your attention Attenuation curve Z = 5 Z = 13 Z = 26 10 1 , cm 2 Z = 82 gr 10 0 10 1 Energy, MeV 17 / 13
Recommend
More recommend
