CEA/Saclay research centre has several 60Co irradiation facilities dedicated to gamma irradiation for both CEA and industrials R&D needs in various fields such as electronuclear, defence, electronics, space as well as health applications.

In the more specific field of health, an irradiator is used to radiosterilise, i.e. to neutralise microbiological contaminants (such as bacteria, viruses, microbes, spores) using ionising radiation, for medical devices such as hip prostheses, orthopaedic screws or plates, on behalf of their suppliers. The great advantage of gamma radiation sterilization, compared with other sterilization alternatives (gas or cold immersion in liquid chemicals), is that medical devices do not have to be removed from their sealed pouches; they are processed directly through their packaging.

Radiosterilization of medical devices is a highly demanding process, in line with the requirements of ISO 13485 and ISO 11137. Firstly, the doses delivered must be neither too low, to ensure product sterility, nor too high, to avoid altering their integrity. Secondly, three qualification stages are required to guarantee validation of irradiation sterilization processes. The first two, known as installation and operational qualification, are respectively designed to demonstrate that the sterilization equipment has been installed in accordance with its specifications and is operating correctly, delivering appropriate doses within the limits of defined acceptance criteria. In particular, operational qualification consists in characterizing the irradiator in terms of dose distribution and reproducibility, by considering the volumes to be irradiated filled with a homogeneous material, with envelope densities representative of the products to be treated. Finally, the third qualification stage, known as performance qualification, must demonstrate, specifically for the medical products to be treated, that the equipment operates consistently, in accordance with predetermined criteria, and delivers doses within the specified dose range, thus producing a product that meets the specified requirement for sterility.


Depending on the supplier, irradiation packaging cartons are generally filled with a variety of different medical products, corresponding to a wide range of sizes and weights. The effects on spatial dose distribution of all possible product loading configurations should therefore be examined, including for different fill rates of the cartons on the irradiator's dynamic conveyor. Finally, it should be noted that the qualification processes must be repeated following any modification likely to affect dose or dose distribution, and therefore each time the sources are restocked. These processes are currently carried out exclusively by measurement, using a multitude of dosimeters appropriately distributed within and on the surface of the packages.

The Laboratoire des Rayonnements Appliqués (DRMP/SPC/LABRA), in charge of the POSEIDON gamma irradiator dedicated to the radiosterilization of medical equipment at CEA/Saclay would like to have a digital tool enabling these validation processes to be carried out by simulation. One of the major benefits expected from the use of this tool is to considerably reduce the downtime of the POSEIDON irradiator, imposed by the experimental validation phases.

The aim of the present thesis is to implement a numerical model of the POSEIDON irradiator, enabling the validation phases to be reproduced by simulation, as quickly as possible, while ensuring the accuracy of the results, to the desired precision. This work will be carried out at the DM2S/SERMA/CP2C laboratory (CEA/Saclay) with regular exchanges with the LABRA laboratory. CP2C is specialized, among other things, in radiation protection studies using numerical simulations.

Thus, the subject of the thesis, divided into three stages, will explore an alternative validation approach to that, carried out experimentally:

• The first stage will involve the development of a numerical model of the POSEIDON irradiator, integrating the dynamic nature of radiosterilization treatments. This numerical model will be based on a calculation methodology to be decided during the thesis (Monte-Carlo or deterministic method), with a compromise between the quality of the results obtained and the speed of calculation execution. For this stage, the radiation transport Monte Carlo code TRIPOLI-4® will be used as a reference, with comparisons made using other numerical tools such as MCNP®, PENELOPE, GEANT4, NARMER-1, etc.;

• The second stage will successively involve validation of the selected numerical model by comparison with experimental measurements, to be defined and carried out during the thesis, and its application to the calculation of operational qualification processes and performances for different families of supplier cartons. As regards validation of the calculations, the instrumentation used for gamma dose measurements will be numerically modelled and analysed, taking into account all the physical phenomena involved in absorbed dose (photon and electron doses). The aim is to consolidate calculation/measurement comparisons for experiments carried out during the thésis;

• The final step will be to analyze the contribution of the numerical model in relation to the experimental approach. This computational approach will nevertheless need to be optimized in terms of calculation time, in order to facilitate the sensitivity analyses to be carried out.

During the thesis, various directions of research will be investigated, such as improving the modelling of reflections during photon transport in a closed environment (PAGURE irradiator casemate; use of deep learning techniques for deterministic codes), implementing stochastic geometries to model the contents of the packaging to be irradiated, and improving algorithms to reduce computation times.