GPU computing of yield stress fluid flows in narrow gaps

IF 2.2 3区工程技术 Q2 MECHANICS

Theoretical and Computational Fluid Dynamics Pub Date : 2023-10-19 DOI:10.1007/s00162-023-00674-x

Ivonne Leonor Medina Lino, Mariana Carrasco-Teja, Ian Frigaard

引用次数: 0

Abstract

We present a Graphic Processing Units (GPU) implementation of non-Newtonian Hele-Shaw flow that models the displacement of Herschel-Bulkley fluids along narrow eccentric annuli. This flow is characteristic of many long-thin flows that require extensive calculation due to an inherent nonlinearity in the constitutive law. A common method of dealing with such flows is via an augmented Lagrangian algorithm, which is often painfully slow. Here we show that such algorithms, although involving slow iterations, can often be accelerated via parallel implementation on GPUs. Indeed, such algorithms explicitly solve the nonlinear aspects only locally on each mesh cell (or node), which makes them ideal candidates for GPUs. Combined with other advances, the optimized GPU implementation takes \(\approx 2.5\%\) of the time of the original algorithm.

Abstract Image

查看原文本刊更多论文

狭缝中屈服应力流体流动的GPU计算

我们提出了一种非牛顿Hele-Shaw流的图形处理单元（GPU）实现，该实现模拟了Herschel-Bulkley流体沿狭窄偏心环空的位移。这种流动是许多细长流动的特征，由于本构定律中固有的非线性，这些流动需要大量计算。处理这种流动的一种常见方法是通过增广拉格朗日算法，该算法通常非常缓慢。在这里，我们展示了这样的算法，尽管涉及缓慢的迭代，但通常可以通过GPU上的并行实现来加速。事实上，这样的算法只在每个网格单元（或节点）上局部地显式地解决非线性方面，这使它们成为GPU的理想候选者。结合其他进展，优化的GPU实现占用了原始算法的时间（约2.5\%\）。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Theoretical and Computational Fluid Dynamics 物理-力学

CiteScore

5.80

自引率

2.90%

发文量

审稿时长

>12 weeks

期刊介绍： Theoretical and Computational Fluid Dynamics provides a forum for the cross fertilization of ideas, tools and techniques across all disciplines in which fluid flow plays a role. The focus is on aspects of fluid dynamics where theory and computation are used to provide insights and data upon which solid physical understanding is revealed. We seek research papers, invited review articles, brief communications, letters and comments addressing flow phenomena of relevance to aeronautical, geophysical, environmental, material, mechanical and life sciences. Papers of a purely algorithmic, experimental or engineering application nature, and papers without significant new physical insights, are outside the scope of this journal. For computational work, authors are responsible for ensuring that any artifacts of discretization and/or implementation are sufficiently controlled such that the numerical results unambiguously support the conclusions drawn. Where appropriate, and to the extent possible, such papers should either include or reference supporting documentation in the form of verification and validation studies.