Year
2007
Abstract
Accurate assay of dense heterogeneous waste forms by high-resolution gamma-ray spectroscopic techniques is made especially challenging by severe and non-homogenous attenuation of the waste material and/or a non-uniform source distribution. The development of the Tomographic Gamma Scanner (TGS) arose out of the desire to address these difficulties. By using first generation computed tomography techniques, an attenuation map of the item is constructed. This attenuation map, combined with passive emission scan data, is used to reconstruct attenuation and activity-location compensated assay values. The approach has been successfully used for nondestructive assay of low to medium density drummed waste. In extending the method to higher densities, one faces the problem of poor counting statistics in both the transmission and emission data. The reconstructed images, which are a prerequisite to forming a representative item-specific matrix correction from the composite data, become diffuse and mottled due to poor precision and are, correspondingly, less valuable. In this work, the formation of the linear attenuation map is specifically examined using an analytical and an experimental approach. The analytical approach examined the sampling frequency of the voxel grid to understand the effects of voxel position on the attenuation map. Experimentally, a set of blocks of various materials was used to construct test patterns closely representing the mathematical model embedded in the TGS image reconstruction software. Results are presented to demonstrate the behavior of contrast resolution and spatial resolution. The dual intensity transmission technique is also discussed as a means for extending the practical density range over which the TGS can be deployed.