Multi-Tokamak Assessment of Modeled Temperature Profiles

Abstract

This study validates the predictive capability of a newly formalized modeling workflow—referred to here as MAESTRO—developed by Dr. Pablo Rodríguez Fernández of the MIT Integrated Modeling Group by comparing simulated plasma temperature profiles with experimental data from three well-documented tokamak discharges: Holland (2011) [1], White (2014) [2], and Zagorski (2015) [3]. The validation study uses the iterative TRANSP and PORTALS transport solvers to achieve flux-matching and self-consistency between heat sources and transport. The experimental temperature and density profiles were used as a starting point for the analysis. Three different SAT rules were used (SAT3 [4], SAT2-EM [5], and SAT2-EM as implemented in ASTRA [6]) and the edge boundary conditions were perturbed ±15% to simulate experimental error. The resulting profiles were plotted against the experimental profiles to validate the model’s accuracy. The percent difference of the simulated and experimental stored energy across the three cases is calculated. The results establish confidence in MAESTRO’s predictive capabilities for predicting future tokamak performance, while identifying areas for model improvement.