The use of a Neutron Generator with the Neutron Back Scattering Method

Transcription

1 The use of a Neutron Generator with the Neutron Back Scattering Method V.R. Bom 11/30/ /30/2005

2 NBS: principle fast neutron slow neutron 3 He tubes source soil mine 2

3 Hydrogen content of some materials Melamine powder RDX TNT NYLON POM (Delrin) Polyethylene Sand (20% water) Sand (10% water) Sand (5% water) Sand (2% water) H-atoms/cm3 1.00E E E E+23 The detector sees much more sand as compared to mine 3

4 Neutron back scattering Advantages speed of detection detect metal free mines deep penetration Disadvantages no identification of materials limitation by soil moisture 4

5 DUNBID 3 He tubes preamplifiers The neutron source may be placed on top of the tubes 5

6 System layout Right hand signals Left hand signals Position module level adapter PC IO board HV 6

7 Specifications device to be dragged over the ground (no standoff ) 3 He tubes: 10 bar, 50 cm x 2.5 cm Ø efficiency: 90% at 25 mev non zero sensitivity to high energy neutrons detector array: 16 tubes, 3 cm pitch resolution: 3 cm x 3 cm 7

8 A first test detector in air, no soil (!?) polyethylene mine + delrin mine, on top of detector tubes always subtraction needed 8

9 Imaging Large field-of-view Additional position information leads to better mine recognition (as compared to point detectors) Mine to no-mine signal ratio high locally: at mine position low for total count rates Data fusion with other imaging techniques But: no imaging!! 9

10 Neutron response 1 3He efficiency 1 Relative detection efficiency Cf spectrum Relative source intensity E-02 1.E+00 1.E+02 1.E+04 1.E+06 1.E+08 Neutron energy (ev) 10

11 t3s 3 standard deviations measurement time only counting statistics Count rate with mine: T, without mine: B, signal-to-noise : f=t/b mine signal: uncertainty, 1 s.d.: t3s( T t3s = 9 B) = 3 T + B ( T B) t( T B) t ( T + B) t3s( T = 9 + B) ( ) B T + 1 B ( ( ) B T 1 9 B ( f + 1) ( f ) B = 11

12 Anti-tank mine AT dummy mine, 4cm x 20 cm Ø, nylon shell, melamine filling 252 Cf source, n/s (0.6 MBq), 2 cm above soil Mine signal at 15 cm: 2028 counts, uncertainty: σ=146 12

13 At the Nuclear Research Center of the Egyptian Atomic Energy Authority With a source PuBe source on top of lid n/s 60 s measuring time 13

14 14 Small AP mine (Cairo)

15 Intensity data reduction: { spectrum(mine) - spectrum(no-mine) } spectrum(no-mine) pixel noise from: soil inhomogeneities: 0.15 counting statistics: s measurement time; source strength: n/s PuBe 15

16 Large anti-tank mine with metal casing (Cairo) 16

17 17

18 NBS with a radioactive source see AT and AP mines quickly Metal and rocks do not give a signal, but wood does Annoying background subtraction Counting statistics only important for very homogeneous soil weak sources or short measurement times Can we get rid of this subtraction operation? Are there ways to reduce the no-mine signal? Use timing information? need a pulsed neutron source: neutron generator 18

19 The right time short time-of-flight directly from source few scatterings, not thermalized inside the mine long time-of-flight drifted through surroundings Such events bear no relation to the mine Add to the background may be excluded using a time window 19

20 Neutron time distribution simulation Optimum timing window for neutron time-of-flight Window depends on mine depth Direct fast neutrons can be removed through timing 20

21 Neutron generator Pulsed neutron generator time effects reduced background vehicle mounted GENIE16C DD neutron generator operated at n/s, 2.5 MeV 1 khz - 3 khz neutron pulse width µs 21

22 50 cm 22

23 Physical layout SOIL need closed lid heavy interference from generator electronics on preamplifiers badly shielded cable (1 khz, 2 kv pulses) awkward geometry neutron emission point off center neutron emission point far from soil ( 10 cm) 23

24 GENIE 16C peculiarities Minimum 5% duty factor restricts minimum neutron pulse width Delay time delay time known only through rough graph unhandy adjustment possible damage Restart current check prohibits restart Cabling EM radiation interference problems 24

25 Neutron pulse (1) 50 µs excitation pulse neutrons come after a delay thermal neutron life time in ms range 25

26 Neutron pulse (2) 26

27 TOF window excitation pulse (65 µs) TOFmax delay TOFmin neutron pulse (50 µs) width TOF window TOFmax, TOFmin=0 TOF window 27

28 TOF time Relative detection eff efficiency TOF 1.E-02 1.E+00 1.E+02 1.E+04 1.E+06 1.E time-of-flight (20 cm, micros) Neutron energy (ev) 28

29 Complex situation neutron pulse distribution moderation absorption in soil energy spectrum TOF: window selects energy energy dependent detection efficiency What does the time window? 29

30 30

31 TOF window delay ( no mine) 31

32 32

33 TOF window width 33

34 Delay 10 µs AP mine 10 cm deep TOF-window influences the mine width 34

35 Mine / no-mine ratio AP mine at 10 cm depth 10 µs delay 35

36 Comparison: source - generator (1) mine/no-mine signal ratio in top of peak source 0.15 generator

37 Comparison: source - generator (2) n-generator applying a TOF window + no subtraction less sensitive for standoff distance variations + improved safety + better mine/no-mine signal ratio in the peak top + smaller mine images expensive and heavy, vehicle mounted Radioactive source + better counting statistics: faster needs no-mine subtraction, but what to subtract very sensitive to standoff distance variations radiological hazard 37

38 A real demining device prime importance insensitivity to standoff avoid no-mine correction high mine/no-mine signal ratio less importance t3s value counting statistics physical layout, weight etc indicates a neutron generator 38

39 Conclusions NBS works in dry soils noise is primarily due to soil inhomogeneities radioactive sources give best statistics TOF method using pulsed n-generators gives the best operational characteristics flat no-mine spectrum insensitivity to standoff distance variations field tests with n-generators/tof still have to be performed 39

40 thank you 40

41 AT-mine with TOF window Neutron generator: n/s, 1 khz repetition rate, 50 µs neutron pulse width, 300 s measuring time, TOF µs AT mine depth: 8 cm t3s = 0.3 s 41

42 AP mine with TOF window Neutron generator: n/s, 2.5 khz repetition rate, 20 µs neutron pulse width, 300 s measuring time, TOF µs AP mine depth: 10 cm 42

43 Summary of t3s values AP mine AT mine Generator Source depth DLM standoff 23 3 off center us n-pulse width us n-pulse width us n-pulse width

44 3 He neutron detection 3 He+n p+t+760 kev E p =570 kev E t =190 kev Total absorption 760 Expected 3 He tube signal distribution Count rate p escape 190 t escape 570 Tube output signal 44

45 45

46 Dummy mine data MINE SIZE SHELLL FILLING COMMENT AT 20x4 nylon Melamine Weight:1120 g DLM2.2 8x5 Lucite Mixture Cape Town University PE 6x4 Polyethylene Massive H-atoms POM 20x4 Delrin Massive H-atoms POM/mel 20x4 Delrin Melamine H-atoms 46

47 Response to hydrogen 1.30 on/off-mine signal ratio Al/melamine POM/melamine POM Nylon PE Hydrogen atoms per mine (10 24 ) Various dummy mines: 4 cm X 6 cm, 3 cm depth. 47

48 Anti-personnel mines Delrin (POM) Polyethylene DLM2 Delrin with melamine 4 AP dummy mines t3s = 23 s, for DLM2 48

49 Sensitivity vs. position DLM2.2 at 5 cm depth, 300 sec, 252 Cf source 70 kn/s 49

50 DLM2.2 dummy mine at the center Total counts: with mine: 5337 without mine: Cf source, n/s Mine signal: 544 Statistical uncertainty: σ = ( ) = t3s = 8.4 s 50

51 Stand-off no standoff: mine-signal=6424 counts, 1σ-uncertainty=320 5 cm standoff: mine-signal=4946 counts, 1σ-uncertainty=420 51

52 Response to waste 52

53 A large anti-tank mine with plastic casing (Cairo) 53

54 54