Speaker
Description
The W7-X stellarator is an experimental device to study the reactor relevance of this concept. Its complex three-dimensional geometry makes full plasma characterization challenging, particularly in the scrape-off layer (SOL) where plasma behaviour can be highly localized. Understanding transport and physics in this region is essential for reactor design. Therefore, a Bayesian inference framework is implemented for W7-X which enables the integrated reconstruction of key plasma parameters such as electron density and temperature from a limited diagnostic coverage.
This contribution presents the first steps in the implementation of the framework and its adaptation to W7-X geometry. Currently the work is focused on specific poloidal planes and so only 2D results will be shown. The framework has been benchmarked by performing inversions of bolometry data with the full Bayesian approach, including Hamiltonian Monte Carlo sampling for uncertainty quantification. The obtained tomograms are compared against previous results from Gaussian Process tomography, with the methods agreeing well with each other. Furthermore, results from adding divertor spectroscopy data as prior constraints to the tomographic inversions are showcased. Also next steps which include preparing for inferring plasma parameters at the divertor site using divertor spectroscopy and the MANTIS [1] imagining system will be highlighted.
[1] A. Perek et al, "MANTIS: A real-time quantitative multispectral imaging system for fusion plasmas.", Rev. Sci. Instrum. 90, https://doi.org/10.1063/1.5115569 (2019)