Abstract

This paper describes the verification and validation (V&V) framework developed for the stochastic Particle-in-Cell, Direct Simulation Monte Carlo code Aleph. An ideal framework for V&V from the viewpoint of the authors is described where a physics problem is defined, and relevant physics models and parameters to the defined problem are assessed and captured in a Phenomena Identification and Ranking Table (PIRT). Numerous V&V examples guided by the PIRT for a simple gas discharge are shown to demonstrate the V&V process applied to a real-world simulation tool with the overall goal to demonstrably increase the confidence in the results for the simulation tool and its predictive capability. Although many examples are provided here to demonstrate elements of the framework, the primary goal of this work is to introduce this framework and not to provide a fully complete implementation, which would be a much longer document. Comparisons and contrasts are made to more usual approaches to V&V, and techniques new to the low-temperature plasma community are introduced. Specific challenges relating to the sufficiency of available data (e.g., cross sections), the limits of ad hoc validation approaches, the additional difficulty of utilizing a stochastic simulation tool, and the extreme cost of formal validation are discussed.

Graphic Abstract

中文翻译：

---

低温等离子体模拟和实验的验证与挑战中的挑战与机遇

摘要

本文介绍了针对随机单元内直接模拟蒙特卡洛代码Aleph开发的验证和确认（V＆V）框架。从作者的角度出发，描述了一种理想的V＆V框架，其中定义了一个物理问题，并在现象识别和排名表（PIRT）中评估和捕获了所定义问题的相关物理模型和参数。展示了PIRT指导的许多V＆V示例，以实现简单的气体排放，以演示应用于实际仿真工具的V＆V过程，其总体目标是明显提高仿真工具结果的可信度及其预测能力。尽管此处提供了许多示例来说明框架的元素，这项工作的主要目的是介绍此框架，而不是提供一个完整的实现，这将是一个更长的文档。对更常见的V＆V方法进行了比较和对比，并介绍了低温等离子体社区的新技术。讨论了与可用数据的充分性（例如，横截面）有关的特定挑战，即席验证方法的局限性，使用随机仿真工具的额外困难以及正式验证的极高成本。

图形摘要