Monte Carlo simulations are a powerful tool to predict the outcome of radiation sterilization. A major advantage is that these simulations can be performed early in the medical device design process when only CAD models of the device exist. Further packaging configurations and sensitivity to variations in placement of the device can be explored virtually whereby enabling a rigorous and reliable design for sterilization process.

Monte Carlo simulations for radiation transport are a well-established tool with various open-source packages available for use. One of the most prominent tools is the Geant4 library, which was originally developed by CERN and is now supported by an international collaboration of scientists and software developers. The downside of such tools is that they do not have native support for importing CAD models and are slow in terms of computation times.

Triple Ring Technologies is developing a simulation tool that provides an easy interface to use CAD models and significantly speeds up the simulation times. In this presentation we show the first results of our new simulation tool, which achieved a significant speed up in simulation times by using modern Graphics Processing Unit (GPU) cards. We simulated the irradiation of a hollow aluminum cylinder with a 1.25 MeV photon beam that modelled a typical gamma sterilization source. Simulations were performed with our new tool and with Geant4. Figure 1 shows the simulated dose profiles along the inner and outer circumferences of the cylinder.

We find excellent agreement between the two simulation engines with the advantage of a 2,500x speed up of our new tool compared to single-threaded Geant4.