Computational Fluid Dynamics on AWS

Computational Fluid Dynamics

on AWS

First Published March 10, 2020

Updated July 27, 2021

This paper has been archived

For the latest technical content, refer t o the HTML

version at:



computational-fluid-dynamics-on-aws/computational-fluiddynamics-on-aws.html

Notices

Customers are responsible for making their own independent assessment of the

information in this document. This document: (a) is for informational purposes only, (b)

represents current AWS product offerings and practices, which are subject to change

without notice, and (c) does not create any commitments or assurances from AWS and

its affiliates, suppliers or licensors. AWS products or services are provided ¡°as is¡±

without warranties, representations, or conditions of any kind, whether express or

implied. The responsibilities and liabilities of AWS to its customers are controlled by

AWS agreements, and this document is not part of, nor does it modify, any agreement

between AWS and its customers.

? 2021, Amazon Web Services, Inc. or its affiliates. All rights reserved.

This paper has been archived

For the latest technical content, refer to the AWS

Whitepapers & Guides page:



Contents

Introduction ..........................................................................................................................1

Why CFD on AWS? .............................................................................................................3

Getting started with AWS ....................................................................................................4

CFD approaches on AWS ...................................................................................................6

Architectures.....................................................................................................................6

Software............................................................................................................................9

Cluster lifecycle ..............................................................................................................12

CFD case scalability .......................................................................................................13

Optimizing HPC components ............................................................................................19

Compute .........................................................................................................................19

Network...........................................................................................................................20

Storage ...........................................................................................................................21

This paper has been archived

Visualization ...................................................................................................................25

Costs ..................................................................................................................................26

For the latest technical content, refer to the AWS

Whitepapers & Guides page:

Contributors .......................................................................................................................

29



Conclusion .........................................................................................................................29

Further reading ..................................................................................................................29

Document revisions ...........................................................................................................30

Abstract

The scalable nature and variable demand of computational fluid dynamics (CFD)

workloads makes them well suited for a cloud computing environment. This whitepaper

describes best practices for running CFD workloads on Amazon Web Services (AWS).

Use this document to learn more about AWS services, and the related quick start tools

that simplify getting started with running CFD cases on AWS.

This paper has been archived

For the latest technical content, refer to the AWS

Whitepapers & Guides page:



Amazon Web Services

Computational Fluid Dynamics on AWS

Introduction

Fluid dynamics is the study of the motion of fluids, usually in the presence of an object.

Typical fluid flows of interest to engineers and scientist include flow in pipes, through

engines, and around objects, such as buildings, automobiles, and airplanes.

Computational fluid dynamics (CFD) is the study of these flows using a numerical

approach. CFD involves the solution of conservation equations (mass, momentum,

energy, and others) in a finite domain.

Many CFD tools are currently available, including specialized and ¡°in-house¡± tools. This

variety is the result of the broad domain of physical problems solved with CFD. There is

not a universal code for all applications, although there are packages that offer great

capabilities. Broad CFD capabilities are available in commercial packages, such as

ANSYS Fluent, Siemens Simcenter STAR-CCM+, Metacomp Technologies CFD++, and

open-source packages, such as OpenFOAM and SU2.

A typical CFD simulation involves the following four steps.

?

Define the geometry ¡ª In some cases, this step is simple, such as modeling

flow in a duct. In other cases, this step involves complex components and

moving parts, such as modeling a gas turbine engine. For many cases, the

geometry creation is extremely time-consuming. The geometry step is graphics

For the

latest atechnical

content,

referpreferably

to the AWS

intensive

and requires

capable graphics

workstation,

with a Graphics

Processing Unit (GPU).

the geometry

is provided

WhitOften,

epapers

& Guides

page:by a designer, but the

CFD engineer must ¡°clean¡± the geometry for input into the flow solver. This can



be a time-consuming step and can sometimes require a large amount of system

memory (RAM) depending on the complexity of the geometry.

?

Generate the mesh or grid ¡ª Mesh generation is a critical step because

computational accuracy is dependent on the size, cell location, and skewness of

the cells. In Figure 1, a hybrid mesh is shown on a slice through an aircraft wing.

Mesh generation can be iterative with the solution, where fixes to the mesh are

driven by an understanding of flow features and gradients in the solution.

Meshing is frequently an interactive process and its elliptical nature generally

requires a substantial amount of memory. Like geometry definition, generating a

single mesh can take hours, days, weeks, and sometimes months. Many mesh

generation codes are still limited to a single node, so general-purpose Amazon

Elastic Compute Cloud (Amazon EC2) instances, such as the M family, or

memory-optimized instances, such as the R family, are often used.

This paper has been archived

1

 ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download