REVOLT-Upgrade: a software tool for fast, energy-resolved simulations of soft x-ray detectors in nuclear fusion experiments

Abstract The design of x-ray diagnostics for nuclear fusion (NF) experiments often requires a simulation tool that balances accuracy, speed, and computational efficiency; especially when several simulations are needed to investigate different experimental conditions, plasma scenarios, optimize diagnostic components and lines of sight configurations and test the feasibility of experimental measurements and analyses. Although Monte Carlo simulation tools (like GEANT4 and MCNP) offer high accuracy, they are often too computationally heavy for NF applications. For this reason, a software tool (written in C and Python) called ‘REVOLT’ was developed, trading a small degree of accuracy for significant gains in simulation speed and resource demands. This paper presents the enhanced capabilities of REVOLT: initially designed to compute transport matrices and shading effects, REVOLT has been upgraded specifically for x-ray pixelated-detector applications on NF experiments. Starting from the emissivity of the x-ray source (for example, a NF plasma), REVOLT-U calculates counts and energy spectra at each detector pixel, accounting for the detector efficiency and the x-ray attenuation through various gases and experimental components. The emissivity of the source can either be derived from independent simulations or directly computed by REVOLT-U using experimental data from different diagnostics. The upgrade is not only useful to improve the simulation of detector data, but also plays a crucial role in enabling tomographic reconstruction algorithms that offer both spatial and energy resolution. This paper focuses on the application of REVOLT-U to NF experiments equipped with gas electron multiplier (GEM) detectors. REVOLT-U has been validated using two different x-ray sources and geometries, with GEANT4 simulations serving as a reference standard. In addition, an example application of REVOLT-U is presented for two NF experiments: RFX-mod2 (Padova, Italy) and MAST-U (Culham, UK). For the MAST-U case, a comparison with experimental data is also included.