Abstract
The interactions of gamma-rays with matter have been studied for many years and there are accurate mathematical representations of the physical processes involved. Tabulations of the total interaction cross-section and of the major constituent processes have been produced which have an uncertainty of a few percent. In recent years the development of the tomographic scanner, first at EMI and then worldwide, has lead to the measurement of material interaction cross-sections with a precision of less than one percent, which is much less than the available tabulations. The form of the tabulations has also meant that data points must be interpolated from standard energy values, and so a large data base must be maintained if the values are stored on a computer. This is a time consuming and fairly inefficient process, especially if a micro-computer is used for the data base. In 1981 a compact, portable computer program was developed for the calculation of total interaction cross-sections (Jackson and Hawkes (1981)), which reproduced the tabulated data with a high precision (typically better than 0.5%) over a wide range of atomic number (1 < Z < 54) and energy (15 < E < 1500 keV). Although the computer code reproduces the theoretical data with high precision, the uncertainty in the original data is relatively large, this has prompted this experimental study to determine the agreement of the theoretical data values with experimental measurements. The precision desired from the experiments at the outset was 1%, which has been achieved in some, although not all, of the results presented here. The range of atomic number and gamma-ray energies of the materials tested in this study are from carbon (Z=6) to lead (Z=82) and from 15 to 1500 keV. This range has been determined largely by the sources and samples that have been available within the department, but it also represents the range of materials and energies that have been used in the industrial tomography group within this department. Isotopic sources and high resolution solid state detectors have been used to measure total cross-sections for the twelve elements used and up to 25 energies per sample (depending on the range of thicknesses available for the material). Many repeat runs have been done on each sample to increase the precision of the measurements presented here. The use of total interaction cross-sections for materials analysis has also been investigated for several situations and these are discussed and it is shown which methodsare possible and which are not with the precision used in these experiments. This study has to be seen as part of an ongoing process of experimentally producing total interaction cross-sections of higher and higher precision for comparison with theoretical formulations so that a deeper understanding of the underlying processes can be obtained. Within that framework some directions for future work have been outlined which will benefit from further study and a more detailed exploration.