The linear attenuation coefficient (μ) quantifies the intensity loss of incident radiation due to absorption and scatter when passing through a specific thickness of an attenuating material, making it vital in radiation-matter interaction studies [1]. The Monte Carlo (MC) simulation method models radiation interaction probabilities with matter, transporting particles, and creating geometries with various materials, and can be validated against experimental data. The PHITS, a MC code, handles nearly any particle's transport across wide energy ranges [2]. Research has shown Monte Carlo codes' efficacy in simulating tissue-equivalent attenuating materials, yielding results compatible with the NIST reference database [3,4]. Previous studies have examined ballistic gel as a potential human muscle tissue surrogate, analyzing its radiological properties. Preliminary results show promising tissue-equivalent characteristics. This work aims to model, simulate, and validate this setup using the Monte Carlo PHITS code (version 3.32) and analyze the μ obtained from experimental results. The geometry of the experimental setup was reproduced and ballistic gel samples measuring 10x10x2 cm³ were inserted between the detector and the source, one at the time (varying from 2 to 8 cm). For the detector, an 3”x3” NaI(Tl) crystal was simulated, following the material specifications [5]. Photons with energies of 186.1, 241.9, 295.2, 351.9, 609.3, 1759, and 2204.1 keV from Ra-226, with corresponding emission probabilities [6]. Both, spectra and monoenergetic conic beams were tested. The experimental setup was shielded with 5 cm lead bricks. The photon beam was collimated using an 8 mm diameter aperture. The T-deposit tally was used to record pulses detected by particles emitted from the source within the sensitive volume of the NaI(Tl). The PHITS code generated an output file with relative errors below 1.7% for the spectrum and 1% for monoenergetic energies, based on a total of 10⁷ particle histories. The linear μ was calculated according to the Attenuation Law equation [7]. The findings showed a high correlation with experimental results, with discrepancies below 5% for all energies except for the lowest energy (186.1 keV), where the difference was 12% for the spectrum. For monoenergetic energies, the difference was < 4% across all energies. These results also indicate that at lower emission probabilities, relative errors have a greater impact. Additionally, the Compton edge contribution from higher energies to the low-energy spectrum and the interaction of the beam with lead, generating photons and electrons, may influence these differences in the spectrum. The obtained values also align with theoretical results from NIST, particularly for monoenergetic beams. In conclusion, the setup was modeled and simulated through PHITS code. For monoenergetic beams, the results exhibited a variation of < 4% when compared to experimental data and < 1% when compared to the NIST reference database. These findings demonstrate the efficacy of the methodology employed and the use of this code for this application.

Acknowledgement: This work was supported by the Centro de Desenvolvimento da Tecnologia Nuclear - CDTN and by Brazilian Institute of Science and Technology for Nuclear Instrumentation and Applications to Industry and Health (INCT/INAIS), CNPq project 406303/2022-3.