Multiple scattering calculations for proton beams: Comparison of results from the general-purpose Monte Carlo codes penh, fluka and topas

Purpose: To test the multiple Coulomb scattering theories implemented in the Monte Carlo simulation codes penh, fluka and topas. Methods: Simulations with the three codes of proton beams with initial energies between 100 and 220 MeV impinging normally on slabs of 14 different compositions and various thicknesses were performed. The simulated angular distributions of transmitted protons are very approximately Gaussian with a characteristic angle ??????0, which measures the spread of the distribution. The characteristic angles resulting from the simulations are compared to experimental data available from two experiments. The degree of agreement is quantified by the ??????2 statistic. The 151 cases considered are analyzed by grouping the data in various ways (all together, by experiment, according to the energy of the proton beam or the slab material). Results: In general, penh produced the better description of the experimental data. If all data are included, the ??????2 values were 3.7 & PLUSMN; 0.6 for penh, 18.7 & PLUSMN; 0.4 for fluka and 7.4 & PLUSMN; 1.5 for topas. The ranges in the values of ??????2 obtained for the various data groups are [0.1 & PLUSMN; 0.2; 10.7 & PLUSMN; 4.1] for penh, [0.10 & PLUSMN; 0.03;221.1 & PLUSMN; 7.8] for fluka, and [0.2 & PLUSMN; 0.3; 46.2 & PLUSMN; 26.4] for topas. The minimum and maximum values in these ranges occur, in the three codes, for the Zn and the brass slabs, respectively, with 158.6 MeV protons. Conclusions: The three codes provide a reasonably accurate description of multiple scattering distributions of protons transmitted through material slabs. Brass is the material whose experimental data are worst described by the three codes analyzed.