IMAS Compatible Neural-Network Accelerated Core-Pedestal Simulations with Self-Consistent Transport of Impurities

STEP (Stability, Transport, Equilibrium, and Pedestal) is a new predictive workflow developed within the OMFIT framework [https://gafusion.github.io/OMFIT-source/] to find stationary plasma scenarios with self-consistent core transport, pedestal structure, current profile, and plasma equilibrium. Key features of the workflow are: (1) Self-consistent modeling of impurity transport reduces the number of free parameters and assumptions that are used in the simulations; (2) Fast yet accurate simulations by leveraging neural network based models for the pedestal structure, neoclassical bootstrap current, and turbulent and neoclassical transport; (3) Full compatibility with the ITER Integrated Modeling and Analysis Suite (IMAS) achieved by transferring information among the different physics components with the newly developed OMAS library [https://gafusion.github.io/omas]. Simulation results and comparison with experimental DIII-D measurements will be presented.