Bremsstrahlung production at 50 MeV in different target materials and configurations

A combination of Monte Carlo, convolution, and experimental techniques have been used to investigate bremsstrahlung production at 50 MeV in full-range targets to produce narrow elementary photon beams for scanning. Calculations using the ITS 3.0 Monte Carlo system for various target designs, including particle transport through the treatment head of an MM5O racetrack microtron and a water phantom, have been compared to experimental dose profiles from narrow photon beams at 10-cm depth in water. A reduction in the ITS 3.0 default substep size has been found necessary even for incomplete agreement, in consistency with the findings of Faddegon and Rogers [Nucl. Instrum. Meth. A 327, 556-565 (1993)] for a different experimental setup and energy using the previous version of ITS. Results show that the calculated shape of the tail of dose distributions from narrow photon beams agrees well with measurements, but CYLTRAN/ITS 3.0 fails to reproduce the central part of the distribution. The discrepancy at small angles, reported previously for EGS4 and ITS 2.1 simulations, possess a limitation to Monte Carlo simulations of narrow photon beams used in scanned systems of clinical accelerators. Radial dose profiles have been calculated by convolution of the energy fluence at the exit of the target with one polyenergetic Monte Carlo calculated dose kernel and also a database consisting of ten different dose kernels corresponding to different monoenergetic photon pencil beams for comparison. The agreement with the much slower fully detailed Monte Carlo calculations was better when using the database kernels than the polyenergetic kernel. Results for the mean energy, mean polar angle, and energy fluence at different depths within various targets have been obtained. These are discussed in the context of the design characteristics of bremsstrahlung targets with emphasis on their utilization for scanning photon beam techniques.