Efficient Parallel Computation of Gamma-Ray Dose Fields Using OpenMP and Its Engineering Applications

This study addresses the limitations of traditional point kernel integration methods in gamma-ray dose field calculations, such as restricted source decomposition capacity, inefficient multi-dose-point synchronization, and high computational costs for large-scale models. We developed a parallel implementation of the point kernel integration method (named Photosh) based on OpenMP, featuring a redesigned source-processing module, a matrix-based parallel computation unit, and dynamic task scheduling.Tests demonstrate that Photosh maintains calculation accuracy while significantly shortening computation time, with a maximum acceleration ratio of 14.29 times. It is particularly suitable for multisource scenarios in nuclear facilities and dense dose field analysis, demonstrating its potential for rapid shielding optimization and accident simulation. This paper also analyzes the deviation characteristics of the program in complex scattering regions and their engineering acceptability, discusses its limitations in high-precision scenarios, and compares its performance with GPU parallel methods and potential for extension.