CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums

Blog Entries

Register Blogs Members List Search Today's Posts Mark Forums Read

Old

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's Avatar
Senior Member
Posted in Uncategorized
Views 565 Comments 0 Hisham is offline Edit Tags
Old

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: 807
Size:  16.3 KB

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

Name:  piso15_wake.jpg
Views: 690
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
Member
Posted in Uncategorized
Views 809 Comments 0 skeptik is offline Edit Tags
Old

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
...
New Member
Posted in Uncategorized
Views 466 Comments 0 beauty is offline Edit Tags
Old

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:	281
Size:	36.3 KB
ID:	120  
Member
Posted in Uncategorized
Views 1003 Comments 4 skeptik is offline Edit Tags
Old

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

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

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: 559
Size:  46.2 KB

The mesh was refined so it was OK after checkMesh

Name:  hybrid_refined.jpg
Views: 576
Size:  83.7 KB

not bad) the next - recalculate LowRe K-eps, just for fun.
Member
Posted in Uncategorized
Views 570 Comments 0 skeptik is offline Edit Tags

All times are GMT -4. The time now is 05:47.