CFD Online Discussion Forums

Floating Point Math in Bash

Posted March 6, 2013 at 14:35 by Hisham (Notes for OpenFOAM!)

This post is copied from: http://www.linuxjournal.com/content/...oint-math-bash
Floating Point Math in Bash, By Mitch Frazier

To use floating point arithmetic in Bash one can source this script (i.e. source floatSupport.sh). The floatSupport.sh is:

Code:

#!/bin/bash 
# 
# Floating point number functions. 

##################################################################### 
# Default scale used by float functions.

...

Hisham

Senior Member

Posted in Uncategorized

Views 1987 Comments 0 Hisham is offline

PISO's strange behavior at different meshes :(

Posted February 25, 2013 at 07:19 by skeptik (CFD notes)

1. mesh quad+tri+3D extruded. at 17 grad most flow around airfoil is laminar

Name: piso_17_hybrid.jpg
Views: 1966
Size: 16.3 KB

2. mesh only tri+3D extruded. at 15 grad flow behind foil is completely turbulent

Name: piso15_wake.jpg
Views: 1594
Size: 20.6 KB

so this pictures shows that we cannot get proper results for Cx, Cy due to bad prediction of the flow detachment.

try to find a solution, may be OF 2.1 will show better results

skeptik

Member

Posted in Uncategorized

Views 2080 Comments 0 skeptik is offline

Bubbly flow boundary conditions

Posted February 20, 2013 at 07:58 by beauty

meeting the same problem

Quote:

Originally Posted by Vivek
;99722

Hey Danny,

It is very simple but not straight forward. In many cases you would seen people model the top gas-liquid surface as outlet using two-fluid model. But, I hope that you would agree with me that two-fluid model is not expected to predict the gas-liquid surface. Ypu can model the top gas-liquid interface as follows:

1. Set top surface as inlet. Set all componet of liquid velocity to zero. Set vertical gas velocity to terminal rise

...

beauty

New Member

Posted in Uncategorized

Views 1291 Comments 0 beauty is offline

Good news about airfoil and OpenFoam

Posted February 20, 2013 at 03:42 by skeptik (CFD notes)

So, at the end of the testing OpenFOAM/simpleFoam, i get:

NACA23012
U=21.3, Re=0.44e6,

LowRe case, LaunderSharmaKE:
1. OF 1.7 cannot solve the case with kEpsilon on hex+tetra 2D mesh. It just crashes.

2. There are a lots of details when you use LowRe LSkEps model, you need lots of iteration, lots of interpolation data to provide better accuracy.

3. People invented k-omega for different flows near the walls. And better one - kOmega...

Attached Thumbnails

Click image for larger version

Name: kEpsilon_lowRe.png
Views: 1197
Size: 36.3 KB
ID: 120

skeptik

Member

Posted in Uncategorized

Views 2511 Comments 4 skeptik is offline

Hey, Dude, be careful with liftDir/dragDir in controlDict

Posted February 19, 2013 at 05:41 by skeptik (CFD notes)

Tags airfoil, openfoam

When U is changed due to angle of attack you should change controlDict in libForces section.

So, what can you get when you have proper controlDict

U=56.4, Re=1.6e6

Name: kOmega_Epsilon.jpg
Views: 1474
Size: 46.2 KB