
[Sponsors] 
July 14, 2009, 01:20 
help in ANDERSON text...

#1 
Senior Member
ATM
Join Date: May 2009
Location: United States
Posts: 104
Rep Power: 9 
i`m going through anderson`s 'CFD and its applications",and i`m stuck on pg. 308
(analysis of quasi one dimensional subsonicsupersonic nozzle). here the initial condition of density,temperature and velocity is given by the equations p = 1  0.3146x T = 1  0.2314x . initial conditions at t = 0 V = (0.1 + 1.09x)T^1/2 where 'x' is the axis where the grid points are evenly spaced forgive me if dis question`s stupid, but what is the origin of these eqns? there is no explanation offered in the text as to what are the governing principles behind these. are these just taken arbitrarily or what... please help. thanks in advance, atm 

July 14, 2009, 09:58 
Initial Conditions

#2 
Senior Member
Richard Smith
Join Date: Mar 2009
Location: Enfield, NH, USA
Posts: 138
Blog Entries: 4
Rep Power: 9 
The subsonicsupersonic nozzle appears in Modern Compressible Flow, 2nd Edition, Anderson, pp. 365 (The TimeMarching Technique) with the following quote:
"...the guessed initial temperature distribution at time t=0. It is arbitrarily taken as linear variation." I presume it's likely the same example in your book "CFD and its Application." So I believe all the initial values are based on knowing the solution, so they have no theoretical foundation. They are probably chosen to be just close enough to the final (known) solution to make your solver work, but not far enough away to cause problems. If you are following along and writing your own solver I would suggest substituting fixed values and see what converged solution you get.
__________________
Symscape, Computational Fluid Dynamics for all 

July 14, 2009, 11:26 

#3 
Senior Member
ATM
Join Date: May 2009
Location: United States
Posts: 104
Rep Power: 9 
thank u very much...
is it better if i write my computer code in C++ or MATLAB? how do the 2 differ? thanks in advance. atm 

July 14, 2009, 14:09 

#4 
Member
Join Date: Mar 2009
Posts: 62
Rep Power: 9 
Since your problem is 1D you can use C++ or Matlab. For 2D and 3D problems Matlab is simply to slow.
Use the language you are more familiar C++ or Matlab for this simple problem. 

July 15, 2009, 01:11 

#5 
Senior Member
ATM
Join Date: May 2009
Location: United States
Posts: 104
Rep Power: 9 
oh... thank u very much.i guess C++ is better.


Thread Tools  
Display Modes  


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
DPM Particle tracks summary text input  Rajesh Gutti  FLUENT  1  October 17, 2008 02:29 
Fluent in text mode  Aris  FLUENT  1  April 15, 2008 07:13 
Saving velocity contours to a text file  Craig Johansen  FLUENT  1  March 6, 2007 20:07 
rsm,Kepsilon models, donating text on turbulence  francisco saldarriaga  CFDWiki  1  May 19, 2006 04:54 
TEXT COMMAND  Alex Munoz  FLUENT  0  April 10, 2003 22:32 