Session: CIE-01-02 AMS General

Paper Number: 116989

116989 - Physics-Informed Multi-Output Surrogate Modeling of Fusion Simulations 

Computational simulation has allowed scientists to explore, observe, and test physical regimes previously thought to be unattainable. High-fidelity models are derived from physical principles and calibrated to experimental data. Validation and uncertainty quantification play crucial roles in extrapolating the use of these physics-based models beyond currently accessible experimental domains. Bayesian analysis provides a natural framework for incorporating the uncertainties that undeniably exist in computational modeling, including missing or unknown physics and experimental and numerical errors. However, the ability to perform quality Bayesian and uncertainty analyses is often limited by the computational expense of first-principles physics models. In the absence of a reliable low-fidelity physics model, phenomenological surrogate or machine learned models can be used to mitigate this expense; however, these data-driven models may not adhere to known physics or properties. Furthermore, the interactions of complex physics in high-fidelity codes lead to dependencies between quantities of interest (QoIs) that are difficult to quantify and capture when individual surrogates are used for each observable. Although this is not always problematic, predicting multiple QoIs with a single surrogate preserves valuable insights regarding the correlated behavior of the target observables and maximizes the information gained from available data. A method of constructing a Gaussian Process (GP) that emulates multiple QoIs simultaneously is presented. As an exemplar, we consider Magnetized Liner Inertial Fusion, a fusion concept that relies on the direct compression of magnetized, laser-heated fuel by a metal liner to achieve thermonuclear ignition. Magneto-hydrodynamics (MHD) codes calculate diagnostics to infer the state of the fuel during experiments, which cannot be measured directly. By providing an understanding of the underlying physical phenomena, MHD codes enable physicists to glean insights into target performance, the impacts of modifications, and sources of degradation within the fuel. The calibration of these diagnostic metrics is complicated by sparse experimental data and the expense of high-fidelity neutron transport models. The use of a surrogate is therefore warranted, the development of which raises long-standing issues in modeling and simulation, including calibration, validation, and uncertainty quantification. The performance of the proposed multi-output GP surrogate model, which preserves correlations between QoIs, is compared to the standard single-output GP for a 1D realization of the MagLIF experiment.

Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525. 

Presenting Author: Kathryn Maupin Sandia National Laboratories

Presenting Author Biography: Kathryn Maupin is a Principal Member of the Technical Staff at Sandia National Laboratories. Her research focuses on model form error quantification and multi-objective surrogate modeling. Broader research interests include Bayesian methods, model validation, sensitivity analysis, and uncertainty quantification.

Authors:

Kathryn Maupin Sandia National Laboratories
Anh Tran Sandia National Laboratories
William Lewis Sandia National Laboratories
Michael Glinsky qiTech Consulting