A Machine Learning Approach to Unsteady Fluid Flow Characteristics in Boiling Approaching Critical Heat Flux

接近临界热通量沸腾时不稳定流体流动特性的机器学习方法

• 批准号: 2657669
• 金额: --
• 依托单位: University of Manchester
• 依托单位国家: 英国
• 项目类别: Studentship
• 财政年份: 2021
• 资助国家: 英国
• 起止时间: 2021 至 无数据
• 项目状态: 未结题
• 来源: https://gtr.ukri.org/projects?ref=studentship-2657669
• 关键词: Machine; Learning; Approach; Unsteady; Fluid

The project looks to discover the reasoning behind particularly good heat transfer abilities of boiling fluids approaching Critical Heat Flux (CHF) in a region known as Departure from Nucleate Boiling (DNB). The student will aim to discover the physical reasoning behind oscillatory behaviour as fluid undergoes DNB to understand how a boiling fluid can be kept in this region for substantial time. Outcomes of this work will provide a working knowledge of the triggers which transform a fluid from the DNB region to CHF and transition boiling beyond. A key facet of this project is determining how best to integrate Machine Learning (ML) technology to advance understanding of physical behaviour and improve speed of computations. Provided the project is successful in this regard, this will ultimately deliver the ability to design systems which can safely operate much closer to CHF than previously. Therefore, this project presents the potential to increase heat transfer through boiling for a wide range of applications. To do so, the student will be systematically developing advanced computational tools to understand the physical behaviour behind each stage of the project. These tools will combine the cutting edge of Computational Fluid Dynamics (CFD) and Machine Learning (ML).

该项目旨在发现在一个被称为偏离核沸腾（DNB）的区域中接近临界热通量（CHF）的沸腾流体的特别好的传热能力背后的原因。学生的目标是发现振荡行为背后的物理推理，因为流体经历DNB，以了解沸腾流体如何在该区域保持相当长的时间。这项工作的结果将提供一个触发器的工作知识，将流体从DNB区域转换为CHF和过渡沸腾。该项目的一个关键方面是确定如何最好地集成机器学习（ML）技术，以促进对物理行为的理解并提高计算速度。如果该项目在这方面取得成功，这将最终提供设计系统的能力，这些系统可以比以前更接近CHF安全运行。因此，该项目提出了通过沸腾增加传热的潜力，用于广泛的应用。为此，学生将系统地开发先进的计算工具，以了解项目每个阶段背后的物理行为。这些工具将联合收割机结合计算流体动力学（CFD）和机器学习（ML）的前沿。