Multidomain and Integrative Capabilities for Large-Scale Systems Simulations with High-Order Methods

这项资助的研究目标是开发新的科学计算算法，涉及多个计算模块的集成，并且在数千个处理器上准确，高效，鲁棒和可扩展。 随着现代计算复杂性的增加，在单个框架内集成多个计算模块变得越来越重要，现代计算往往涉及大规模和/或多规模系统。 该项目将专注于开发高阶谱元流体动力学和传热求解器中的多域功能，并确保耦合算法的高阶全局精度和稳定性。 多域能力将包括重叠和移动网格系统，并可应用于多尺度和多物理性质的许多重要问题。 算法和计算的发展将与已知的分析解决方案和现有的实验数据进行系统验证，如果成功，这一努力将产生一个准确和可扩展的计算工具，配备集成功能，使科学计算在21世纪世纪，可作为开源软件在公共许可证下提供。 新工具将允许多学科性质的重要问题的有效计算分析。 一个这样的问题是旋转风力涡轮机的性能受到复杂的大气流入，这可以更准确地预测和优化与开发的方法的可用性。 现代计算的进步和能力以及它们在解决现实工程问题方面的重要性将成为多个推广和教育活动的中心主题，包括妇女通过亚利桑那州立大学的女孩工程项目参与科学，与滕佩的Kyrene学区的K-12学生和教师合作，并组织国家实验室合作者的校园访问，使学生了解该领域的最新进展。

The research objective of this grant is to develop new algorithms for scientific computations that involve integration of multiple computational modules and are accurate, efficient, robust, and scalable on thousands of processors. Integration of multiple computational modules within a single framework becomes increasingly important with the rise in complexity of modern computations that tend to involve large-scale and/or multi-scale systems. This project will concentrate on developing multi-domain capabilities within the high-order spectral-element fluid dynamics and heat transfer solver and ensuring high-order global accuracy and stability of the coupled algorithm. The multi-domain capabilities will include overlapping and moving grid systems and can be applied to many important problems of multi-scale and multi-physics nature. The algorithmic and computational developments will be systematically validated versus known analytical solutions and existing experimental data.If successful, this effort will produce an accurate and scalable computational tool equipped with integrative capabilities to enable scientific computing in the 21st century, available as open source software under a public license. The new tool will allow for efficient computational analysis of important problems of multidisciplinary nature. One such problem is the performance of rotating wind turbines subject to complex atmospheric inflow, which can be more accurately predicted and optimized with the availability of the developed methods. Modern computing advances and capabilities as well as their importance in solving realistic engineering problems will be a central theme for multiple outreach and educational activities, including participation of women in science trough Girls in Engineering Program at ASU, working with K-12 students and teachers in the Kyrene School District in Tempe, and organizing on-campus visits of national lab collaborators to expose students to the most recent advances in the field.