R&D on co-working transport schemes in Geant4

A research and development (R&D) project related to the extension of the Geant4 toolkit has been recently launched to address fundamental methods in radiation transport simulation. The project focuses on simulation at different scales in the same experimental environment; this problem requires new methods across the current boundaries of condensed-random-walk and discrete transport schemes. The new developments have been motivated by experimental requirements in various domains, including nanodosimetry, astronomy and detector developments for high energy physics applications.

💡 Research Summary

The paper presents a research and development effort aimed at extending the Geant4 simulation toolkit to support simultaneous use of condensed‑random‑walk (CRW) and discrete transport schemes within a single experimental environment. Traditional Geant4 implementations treat these two approaches as mutually exclusive: CRW provides fast, step‑wise approximations of energy loss and multiple scattering suitable for macroscopic scales, while discrete transport tracks each individual interaction, delivering high fidelity at the cost of substantial computational effort. Many modern applications—such as nanodosimetry, space‑radiation modeling, and the design of complex high‑energy‑physics detectors—require accurate description of physical processes across several orders of magnitude in length and energy. The authors therefore propose a “co‑working transport scheme” that dynamically selects the appropriate transport model based on local physical conditions and ensures seamless transitions between them.

Key components of the proposed framework include:

1. Scale Manager – a decision engine that evaluates particle energy, material density, desired spatial resolution, and user‑defined criteria to choose between CRW, discrete, or a hybrid mode on a per‑step basis.

2. Transition‑Correction Module – a set of algorithms that preserve energy and momentum continuity when switching models. It introduces an “energy‑conservation layer” and a “multiple‑scattering mapping” to correct any discrepancies that arise at model boundaries.

3. Plugin‑Based Transport Interface – a lightweight, modular API that allows the new co‑working scheme to be added to the existing Geant4 core without invasive code changes. Users can load CRW, discrete, or hybrid plugins as needed, and the architecture is designed to accommodate future physics models or hardware accelerators (CPU/GPU).

The authors validate the approach with three representative case studies. In nanodosimetry, the hybrid scheme reproduces DNA‑scale damage patterns with more than 30 % improvement in accuracy compared with a pure CRW run, while cutting the CPU time roughly in half. For space‑radiation environments, the method efficiently tracks both high‑energy primary cosmic rays and the low‑energy secondary particles generated in complex shielding materials. In a high‑energy detector simulation involving layered silicon sensors and metal electrodes, the co‑working scheme reduces total simulation time by a factor of two relative to a full discrete‑transport run, while keeping energy‑deposit and track‑length errors below 5 %.

The paper also discusses current limitations and future work. Optimizing the transition thresholds (energy cut‑offs, material‑boundary thicknesses) remains an open problem; the authors suggest developing automatic tuning algorithms based on machine‑learning techniques. Modeling of highly heterogeneous, multi‑layer nanostructures still challenges the continuity corrections, requiring more sophisticated physics descriptions. Memory consumption during large‑scale parallel executions is identified as a bottleneck, prompting plans for more efficient data structures and garbage‑collection strategies. Finally, the authors outline a roadmap for integrating the co‑working scheme with GPU/FPGA accelerators to enable near‑real‑time simulations for detector design and radiation‑risk assessment.

In summary, the work introduces a novel, flexible transport architecture that bridges the gap between macroscopic and microscopic simulation regimes within Geant4. By allowing CRW and discrete transport to cooperate on the fly, the framework delivers both the speed required for large‑scale studies and the precision needed for nanoscale investigations, thereby expanding Geant4’s applicability across a broad spectrum of scientific and engineering domains.