Monte Carlo Simulation for Radiation Tomography

Robin P Gardner

Center for Engineering Applications of Radioisotopes

Box 7909, Nuclear Engineering Department, North Carolina State University, Raleigh, NC 27695-7909 gardner@ncsu.edu

ABSTRACT

The general purpose Monte Carlo codes like MCNP are becoming more and more general and, therefore, they can be used for more and more problems in radiation transport. However, radiation transport problems of interest are becoming more and more specialized and complex - so there are still cases where one would want to devise and use a specific purpose Monte Carlo code. I believe this is the case for Monte Carlo simulation of (industrial) radiation tomography.

There are a number of reasons for wanting to devise a specific purpose Monte Carlo code for radiation tomography. Traditionally the general purpose Monte Carlo codes have had problems treating radiation detection spectroscopy and other radiation detection specialized cases like coincidence and anti-coincidence counting. While some of the problems, like simultaneous use of both physical and arti_cial splitting in spectroscopy, are being fixed; the problem remains that some radiation detection mechanisms are still largely unknown and one must resort to semi-empirical methods (detector response functions) to obtain adequate accuracy.

Another reason for using specific purpose codes for radiation tomography is that there are relatively simple variance reduction techniques that can be used in this application. This includes the principle of "forcing", ray tracing and stratified sampling, and expected value splitting. Examples of the use of these methods in specific purpose Monte Carlo codes for specialized tomography applications are presented and discussed in this presentation. This includes: (1) a combined transmission and scatter tomographic system, (2) use of "Ross" filters with X-ray microfocus systems, (3) a Compton camera system, and (4) prompt gamma-ray systems.