Abstract Details
Abstracts
Author: Preeti Sar
Requested Type: Poster
Submitted: 2026-03-20 13:10:33
Co-authors: S.D. Pascuale, H. Dudding, G. Staebler
Contact Info:
Oak Ridge National Laboratory
1 Bethel Valley Road
Oak Ridge, Tennessee 37830
USA
Abstract Text:
Developing a machine learning framework for accurately determining transport fluxes using the quasilinear TGLF model
Preeti Sar1, Sebastian De Pascuale1, Harry Dudding2, Gary Staebler1
1 Oak Ridge National Laboratory, Oak Ridge, Tennessee, USA
2 United Kingdom Atomic Energy Authority, Abingdon, Oxfordshire, United Kingdom
A new neural network model for a quasilinear saturation rule has been developed to map linear gyrokinetic data to nonlinear saturated potential magnitudes to predict the total energy and particle fluxes. The training dataset used is taken from the high-resolution simulation database generated from nonlinear CGYRO for developing the SAT3 model. Overall, SAT3-NN is able to capture the 1D squared potential magnitudes of the dataset more accurately than SAT3, as depicted by lower percentage errors in the peak locations and peak values of the 1D squared potentials. The resulting fluxes also had smaller deviations from the nonlinear CGYRO data as compared to previous saturation models such as SAT0 - SAT2. Consistent with SAT3, SAT3-NN is able to recreate the turbulent flux characteristics such as anti-gyroBohm scaling of fluxes seen for TEM-dominated cases. We further implement integrated modeling to assess the uncertainty of the neural network in practical applications.
Acknowledgement
This work is supported in part by Contract No. DE-AC05-00OR22725 at Oak Ridge National Laboratory, which is managed by UT-Battelle, LLC for the Office of Science of the U. S. Department of Energy.
Characterization: 6.0
Comments: