Year
2013
Abstract
Transmission imaging relies on comparing the transmission with the object present (I) to that when the object is absent (I0), using the relationship I/I0 = exp[-µx], where µ is the attenuation coefficient and x is the path length through the object. Transmission imaging is difficult when a stationary object is adjacent to a wall since the imaging detectors cannot be positioned behind the object for a transmission measurement. One possible solution is to locate the imaging detectors behind the wall. Measurement of the wall’s characteristics can be accomplished by two different methods. In the first method, the imaging system is raised above the object and I0 is measured through the wall (assumes the wall composition does not vary with height). The I measurement is then taken through both the object and the wall. Since the wall was included in the I0 measurement, I/I0 will have the wall effects removed from the transmission image. The second method involves performing a traditional I0 measurement (without the wall). In this method, I/I0 = exp[-µwxw]exp[-µx], where µw is the attenuation coefficient of the wall material and xw is the path length through the wall. The wall characteristics are then estimated with a parametric fitting algorithm. This paper presents a fast neutron imaging measurement of a depleted uranium casting that was performed using concrete blocks to simulate a wall. The Thermo Fisher API- 120 DT neutron generator was located in front of the casting, and the imaging detectors were located behind the concrete block wall. The wall effects are estimated using the two methods previously described. Using the first method, the object’s thickness and attenuation coefficient are determined from measured data within 12% and 14%, respectively. The second method (not including the wall in the I0) performed much better when the thickness of the wall was known and included as an input to the parametric fitting model, correctly yielding the object’s thickness and attenuation coefficient within 4% and 1%, respectively. This experiment demonstrates how a fast neutron image can be obtained for a stationary object near a wall.