SUMMARY

A Varian 2100C 18 MV photon beam has been modeled in this work using the MCNP5 Monte Carlo particle transport user code. The subsequent beam irradiation was also delivered to a water phantom and benchmarked against experimentally measured depth dose data. The model presented in this work establishes the foundation to which further beam characteristics tuning is required in order to realistically model the beam mentioned above. It has been determined in this work that the initial electron beam energy of this beam model is sufficiently close to the electron beam energy from the linear accelerator used to obtain the benchmark depth dose data. Software was developed in this work that divides and configures the treatment head of a medical linear accelerator into several stages and generates the necessary MCNP input files that simulate the beam delivery based on user input. An 18 MV beam was created and collimated to a 4 cm by 4 cm square field size. This collimated beam was then delivered to a water phantom with 0.15 cm length cubic *F8 tally voxels spanning half of an in-plane profile from the central axis. Simulation times for the 3 stages of this configuration took approximately 7.3 hours, 8.4 hours, and 3.2 weeks for stages 1, 2, and 3 respectively. Percentage depth dose values derived from average absorbed dose estimations along the central axis were found to be within ±2% at depths beyond the point of maximum dose deposition. Although the average values are accurate along the cax, the experimental error is undesirably high in some regions of the in-plane profile. The lateral absorbed dose estimations were higher toward the edges where the stereotypical 'horns' exist for higher energy megavoltage photon beams relative to the beam benchmark data.