CDS&E: Collaborative Research: Fast Numerical Simulations of Low Void Fraction Disperse Multiphase Systems using Event-Triggered Communication

CDS

基本信息

批准号：
1953090
负责人：
Vijay Gupta
金额：
$ 27万
依托单位：
University of Notre Dame
依托单位国家：
美国
项目类别：
Standard Grant
财政年份：
2020
资助国家：
美国
起止时间：
2020-06-01 至 2022-05-31
项目状态：
已结题

来源：
https://www.nsf.gov/awardsearch/showAward?AWD_ID=1953090&HistoricalAwards=false
关键词：
CDS&amp Collaborative Research Fast Numerical

项目摘要

Disperse flows involve multiphase fluids where one phase consists of particles, including bubbles and drops, suspended in another contiguous phase. Disperse flows arise in many problems of practical interest in engineering and the natural environment. Direct numerical simulations (DNS) have proven to be indispensable in understanding such flows. However, for flows with more concentrated particles (i.e. low void fractions), it becomes computationally expensive to simulate interactions between particles. Even state-of-the-art simulations take much too long for them to become commonplace. This project will develop new computational methods, based on ideas from control theory, to speed up such simulations and use them to conduct an in-depth study of disperse multiphase systems with low void fractions. The primary intellectual merit of the project lies in using computation algorithms that reduce communication among multiple processing elements for conducting large-scale simulations of heterogeneous multiphase systems - in particular, low void fraction disperse systems where the inclusion of many particles requires large computational domains and long computational times to collect converged statistical quantities. Such averaged data is critical for the construction of reduced order models used for industrial predictions. This project will aid in better understanding of multiphase flows and systems that are critical in energy conversion, material processing, the chemical industry, atmospheric processes, and living systems, as well as develop novel computational strategies based on event-triggered communications between processors to improve the efficiency of parallel computations. Further, people will be educated in both their development and use, and educational material aimed at lowering the barrier to entry for new researchers will be developed.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

分散流涉及多相流体，其中一相由悬浮在另一连续相中的颗粒（包括气泡和液滴）组成。分散流出现在工程和自然环境中的许多实际问题中。事实证明，直接数值模拟（DNS）对于理解此类流动是不可或缺的。然而，对于具有更集中颗粒（即低空隙率）的流动，模拟颗粒之间的相互作用在计算上变得昂贵。即使是最先进的模拟也需要很长时间才能变得普遍。该项目将基于控制理论的思想开发新的计算方法，以加速此类模拟，并利用它们对低空隙率的分散多相系统进行深入研究。该项目的主要智力优点在于使用计算算法来减少多个处理元件之间的通信，以进行异质多相系统的大规模模拟 - 特别是低空隙率分散系统，其中包含许多颗粒需要大的计算域和长的计算域。收集收敛统计量的计算时间。这种平均数据对于构建用于工业预测的降阶模型至关重要。该项目将有助于更好地理解在能量转换、材料加工、化学工业、大气过程和生命系统中至关重要的多相流和系统，并开发基于处理器之间事件触发通信的新颖计算策略，以改进并行计算的效率。此外，人们将接受关于其开发和使用的教育，并将开发旨在降低新研究人员进入门槛的教育材料。该奖项反映了 NSF 的法定使命，并通过使用基金会的智力价值进行评估，被认为值得支持以及更广泛的影响审查标准。