Luk O O, Lakhlili J, Hoenen O, von Toussaint U, Scott B D, Coster D P
Max-Planck-Institut für Plasmaphysik, Garching, Germany.
Philos Trans A Math Phys Eng Sci. 2021 May 17;379(2197):20200074. doi: 10.1098/rsta.2020.0074. Epub 2021 Mar 29.
Harnessing energy produced by thermonuclear fusion reactions has the potential to provide a clean and inexpensive source of energy to Earth. However, throughout the past seven decades, physicists learned that creating our very own fusion energy source is very difficult to achieve. We constructed a component-based, multiscale fusion workflow to model fusion plasma inside the core of a tokamak device. To ensure the simulation results agree with experimental values, the model needs to undergo the process of verification, validation and uncertainty quantification (VVUQ). This paper will go over the VVUQ work carried out in the multiscale fusion workflow (MFW), with the help of the EasyVVUQ software library developed by the VECMA project. In particular, similarity of distributions from simulation and experiment is explored as a validation metric. Such initial validation measures provide insights into the accuracy of the simulation results. This article is part of the theme issue 'Reliability and reproducibility in computational science: implementing verification, validation and uncertainty quantification '.
利用热核聚变反应产生的能量有潜力为地球提供清洁且廉价的能源。然而,在过去的七十年里,物理学家们认识到创建我们自己的聚变能源非常难以实现。我们构建了一个基于组件的多尺度聚变工作流程,用于对托卡马克装置核心内的聚变等离子体进行建模。为确保模拟结果与实验值相符,该模型需要经过验证、确认和不确定性量化(VVUQ)的过程。本文将借助VECMA项目开发的EasyVVUQ软件库,介绍在多尺度聚变工作流程(MFW)中开展的VVUQ工作。特别是,探索模拟和实验分布的相似性作为一种验证指标。这种初步的验证措施有助于深入了解模拟结果的准确性。本文是主题为“计算科学中的可靠性和可重复性:实施验证、确认和不确定性量化”的一部分。