Description
Reliable modelling of pellet fuelling is a key ingredient for integrated modelling in present fusion devices and in ITER-relevant regimes. This work presents a significant upgrade to the pellet modelling code HPI2 [1,2]. Traditionally, HPI2 relied on a constant space discretization parameter for calculating plasmoid properties within the plasma, introducing a degree of arbitrariness and limiting robustness across different injection conditions. This study introduces a dynamic discretization scheme in which the step parameter is computed multiplying the pellet exit time (from its ionized cloud) by its velocity. By directly linking the discretization to the underlying pellet ablation physics, the new mechanism removes the need for ad-hoc tuning, improves numerical stability, and ensures consistent predictive performance across a wide range of pellet injection scenarios. This enhancement strengthens the reliability of HPI2 for integrated modelling applications.
Furthermore, for the first time, extensive validation of the upgraded HPI2 code was conducted using the WEST tokamak. This validation was conducted comparing the results of HPI2 with a pellet fuelled WEST ohmic discharge, first, running HPI2 stand-alone and comparing the line integrated density measurements from interferometry. Then, the updated HPI2 was integrated within the High Fidelity Plasma Simulator (Python workflow originating from JINTRAC) using IMAS I/O, and a full radius predictive modelling study of the full discharge flat top was conducted, coupling various models to HPI2 to predict particle and heating transport, current diffusion, and impurities (NCLASS, TGLF SAT2, SANCO, etc.).
The upgraded code demonstrated good agreement with experimental measurements, in both stand-alone an integrated modelling exercises, falling within expected margins of error for pellet mass, velocity and measurements uncertainties. Consequently demonstrating the capabilities of integrated modelling in pellet scenarios in WEST.
[1] B. Pegourié et al., Nuclear Fusion, 47, 44 (2006)
[2] F. Koechl et al., EUROfusion Preprint EFDA-JET-PR (12), 57, (2012)