IAEA Technical Meeting on Uncertainty Assessment and Benchmark Experiments for Atomic and Molecular Data for Fusion Application

Ballance, C. (Advisor)

School of Mathematics and Physics

Activity: Talk or presentation types › Invited talk

Description

Computer simulations play a critical role in understanding physical and chemical systems. In the last decades verification, validation and uncertainty quantification (VVUQ) of computer simulations has become an important research topic in the physical and chemical sciences and engineering. In support of optimization-based design of a fusion reactor and high precision diagnostics of plasmas in fusion devices the IAEA Atomic and Molecular (A+M) Data Unit has been pursuing VVUQ activities for theoretical A+M data. Our objective is to encourage and support the data development community to provide data users with reasonable uncertainty estimates for calculated A+M data and especially for calculated cross sections. After a series of meetings and discussions with theoretical atomic physicists organized by the Unit it has become clear that the VVUQ methods for theoretical data should develop in collaboration with experimental physicists and benchmark experiments.

Period	09 Dec 2016 → 10 Dec 2016
Held at	The United Nations
Degree of Recognition	International

Documents & Links

The meeting is meant to advance data-oriented research on A+M processes and properties that are important for plasma simulations and plasma diagnostics for fusion energy development and other plasma applications. Leading theoretical and experimental A+M physicists are invited along with plasma physicists using A+M data to discuss uncertainty assessments and benchmark experiments for A+M data for fusion and other plasma applications. The discussion topics are as follows: Prioritization of A+M data needs for fusion applications. Experimental methods and systems to benchmark theoretical A+M collision data. Target uncertainties of A+M collision data for fusion applications. Uncertainty quantification methods for theoretical A+M data. Network of A+M physicists doing benchmark experiments and computations. Among other things the meeting will identify the most important data needs for fusion applications, decide what can be done to benchmark those data and formulate a plan of action among experimental and theoretical A+M physicists to carry out this benchmarking and perform the associated uncertainty assessment. Questions to be addressed include: How can benchmarking measurements be organized taking into account the most recent developments in the experimental as well as theoretical areas? How can detailed measurements of the reaction dynamics, as benchmarks for theory, help to improve reliability of theoretical calculations for a given collision system? The discussions should lead to specific proposals for measurements to set new benchmarks for prioritized data. These measurements will obviously be very difficult in some cases, but they will set the standard. Along with the discussions about new benchmark experiments there will be discussions and proposals about suitable theoretical methods to compute these data, and how and by whom this can be done. The calculations should be accompanied by a full discussion of the inherent uncertainties