
[Sponsors] 
March 10, 2010, 18:52 

#21 
Senior Member
karine
Join Date: Nov 2009
Posts: 158
Rep Power: 8 
U can do the mean values on Matlab (or excel)
If u want to have the means all over the domain, iterate again in Fluent for 1 period of time and dont forget to put yes for data sampling in the iterate menu 

March 10, 2010, 19:01 

#22 
Member
Behrang Asgharian
Join Date: Oct 2009
Posts: 38
Rep Power: 8 
Hi
Thanks. The matlab or excel choice work good. After checking the data sampling in the iterate menu. Where exactly should we go to after the iteration is over? 

March 10, 2010, 21:13 

#23 
Senior Member
karine
Join Date: Nov 2009
Posts: 158
Rep Power: 8 
AFter u go to plotxy or display contours menus. U choose Unsteady statistics .
( if ure flow is stationnary, i think it is better for u to do the means values over 45 periods....) 

March 10, 2010, 21:26 

#24 
Member
Behrang Asgharian
Join Date: Oct 2009
Posts: 38
Rep Power: 8 
Thanks a lot


March 24, 2010, 19:34 

#25 
Member
Behrang Asgharian
Join Date: Oct 2009
Posts: 38
Rep Power: 8 
Hi Again!
How can I initialize the domain using UDF? I mean, for example: I am modeling acoustic oscillating flow which is inherently unsteady so I need to initialize it and after some time it produces sinusoidal curves for velocity and pressure. the problem is that I have to wait for the case to reach that quasi steady state (I mean oscillating with constant frequency and constant amplitudes). If there is a way to initialize with the theoretical sinusoidal pressure distribution throughout the domain, it will definitely reach the quasi steady state faster and that's timesaving. Any suggestions? Thanks 

March 24, 2010, 20:40 

#26 
Senior Member
Micael Boulet
Join Date: Mar 2009
Location: Quebec, Canada
Posts: 112
Rep Power: 10 
Hi Arash,
I have a little comment about your mesh. You can improve it by using unstructured tetrahedral cells. The boundary around the cylinder can be regular, though, thanks to boundary layers utility in Gambit.  Edit: sorry, I mean unstructured quadrilateral Last edited by Micael; March 26, 2010 at 09:12. 

March 25, 2010, 16:55 

#27 
Senior Member
karine
Join Date: Nov 2009
Posts: 158
Rep Power: 8 
Hi
i dont know how to do this...sorry i can suggest that u use large time steps so u can get a quick solution After getting this 'bad' solution, u can use smaller time steps 

March 25, 2010, 17:34 

#28 
Member
Behrang Asgharian
Join Date: Oct 2009
Posts: 38
Rep Power: 8 
Hey!
Thanks for taking the time to answer this. I did that once. The thing is when i use big time steps it gets steady but the amplitudes are not correct and when I change delta t to a small one then, it gets unsteady again and takes some time to become steady. I think the amplitude problem with big delta t's is due to numerical diffusion. Do you think so? Thanks Behrang 

March 25, 2010, 19:44 

#29 
Senior Member
karine
Join Date: Nov 2009
Posts: 158
Rep Power: 8 
Hi
In fact , u will have a bigger numerical diffusion ( truncation error) and u will miss some of the flow motion. If u use a time step dt, u cant correctly simulate any motion with a time charachteristic less then dt/2 (shannon's theorem) ( U may be also violating the stability limits of ure scheme, what will give rise to wiggles) Regards 

June 6, 2010, 23:29 
Problem with flow rate enthalpy sign

#30 
Member
Behrang Asgharian
Join Date: Oct 2009
Posts: 38
Rep Power: 8 
I believe you can help me with this.
I put some Lines (from the menu 'surfaceline') on my model in order to monitor the enthalpy flow rate. I use the flow rate integral which is integral of rho*h*u.dA the absolute values are OK but the sign is negative when it must be positive. I think this has got something to do with the inner product of u.dA. Does anybody know how I should define the monitor line in order to get positive enthalpy flow rates? By the way, if I use 'surfacerake' instead of 'surfaceline', what will happen? Thanks Behrang 

June 16, 2010, 19:40 
Saving a particular profile and load it in another case

#31 
Member
Behrang Asgharian
Join Date: Oct 2009
Posts: 38
Rep Power: 8 
Hi
I am running a case for which I need to give Fluent an initial condition. My question is: Can I save for example density throughout a case and load it in another case with different boundary conditions but with the same mesh? You know, I don't want all the data. I just need a part of it. Is this possible in Fluent??? Thanks Behrang 

Thread Tools  
Display Modes  


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
benchmark: flow over a circular cylinder  goodegg  Main CFD Forum  12  January 22, 2013 12:47 
Moving a Cylinder in cross flow after solving Dynamics equations of motion  maruthamuthu_venkatraman  OpenFOAM  1  November 19, 2009 14:55 
Flow over a flat plate & Flow over a cylinder  cfdxue  Main CFD Forum  0  November 27, 2007 00:26 
Anybody knows smth abt the flow around a cylinder?  Wenxuan  Main CFD Forum  3  March 20, 2007 17:19 
Flow over a cylinder  Anna  Main CFD Forum  9  March 24, 2006 15:32 