We report on the characterization efforts on a largescale, coded-aperture imager for fast neutrons at the U.S. Naval Research Laboratory. The fast neutron imager discussed in this study is designed for a mobile unit (20 ft. ISO container) and consists of a 64-element array (arranged in an 8 by 8 grid pattern) of 15 cm × 15 cm × 15 cm liquid scintillation detectors (EJ-309) mounted behind a 12 × 12 pseudorandom coded aperture. The coded aperture is composed of 15 cm × 15 cm × 10 cm blocks of high-density polyethylene (HDPE). The 12 × 12 pseudorandom coded aperture produces a shadow pattern on the detector array behind the mask. With knowledge of the mask pattern and the number of neutrons detected in a masked and unmasked detector, a source image can be deconvolved to obtain a 2-d image. The number of neutrons detected in each detector was found by processing the fast signal from each photomultiplier tube (PMT) in flash digitizing electronics. The γ-ray background was filtered out using digital pulse shape discrimination (PSD). The 64-element instrument was tested at an indoor facility using four γeutron sources: 1.3-μCi Cf (5.6 × 10 neutrons/s), 9.3-μCi Cf (4.0 × 10 neutrons/s), 99.8-μCi Cf (4.3 × 10 neutrons/s), and a 46-μCi AmBe (2.0 × 10 neutrons/s) at three standoff distances of 9, 15, and 26 m (maximum allowed in the facility) over various integration times. The source was moved at one pixel increments to test the imaging capabilities of the instrument. We found the detection significance at each of these distances and pixel locations with the three sources after 5 min to 25 min of integration depending on the strength and location of the source. We show the detection significance as a function of time at some locations and compare the detected significance to our group's previous 32-element array detected significances [1]. We also shielded Cf sources with 2.5 cm, 5.1 cm, and 10.2 cm thick HDPE and show the detection significance as a function of time at 9 m, 15 m, and 26 m.
展开▼
