
[Sponsors] 
March 24, 2012, 03:04 
how to obtain divergence via FVM

#1 
Member
jack
Join Date: Jul 2011
Posts: 52
Rep Power: 7 
hello everyone
I need to calculate a form like div (coef * grad Phi), I think it is better to assemble it as a summation of fluxes around a control volume . Can you please tell me how to implement it via fortran?Thank you very much! Regards! lg88 

March 25, 2012, 06:43 

#2 
Super Moderator
Glenn Horrocks
Join Date: Mar 2009
Location: Sydney, Australia
Posts: 12,831
Rep Power: 100 
What is Phi? Is the coefficient a constant?


March 26, 2012, 00:05 

#3 
Member
jack
Join Date: Jul 2011
Posts: 52
Rep Power: 7 
Phi is a additional variable and the coef is a constant.Thank you for your help.
Regards lg88 

March 26, 2012, 00:51 

#4 
Super Moderator
Glenn Horrocks
Join Date: Mar 2009
Location: Sydney, Australia
Posts: 12,831
Rep Power: 100 
You might be able to do this without fortran. If you define an algebraic additional variable as phi.grad, then take the div of that algebraic additional variable you might be able to do it entirely in CEL. Then you don't need to worry about the details of summing fluxes and all that.


March 26, 2012, 01:26 

#5 
Member
jack
Join Date: Jul 2011
Posts: 52
Rep Power: 7 
Thanks to ghorrocks.After I get the phi.grad ,how can I take the div of that algebraic additional variable?Are there any use like .grad?
Regards! lg88 

March 26, 2012, 01:29 

#6 
Super Moderator
Glenn Horrocks
Join Date: Mar 2009
Location: Sydney, Australia
Posts: 12,831
Rep Power: 100 
There is not a function for div like that, so you just write the equation out in its direction components long hand. So you define one equation for each cartension direction.


March 26, 2012, 01:44 

#7 
Member
jack
Join Date: Jul 2011
Posts: 52
Rep Power: 7 
So I need to calculate the gradient of the phi.grad in x,y,z direction respectively and then add them together ?
I have another question.I need to use the div(coff*grad phi) scale with the volume of the cell.Are there any word standing the volume of the cell, just like the u stand the velocity of x direction? regards! lg88 

March 26, 2012, 01:54 

#8 
Super Moderator
Glenn Horrocks
Join Date: Mar 2009
Location: Sydney, Australia
Posts: 12,831
Rep Power: 100 
Have a look in the list of available variables in the CFX reference guide.
Also, don't forget CEL has functions like volumeInt() so if you use them you might not need to get the volume either. 

March 26, 2012, 02:39 

#9 
Member
jack
Join Date: Jul 2011
Posts: 52
Rep Power: 7 
oh,I find that it is not easy as you said.
Firstly, ''define an algebraic additional variable as phi.grad".Is it a vector? then where can I write the expression .grad?And when I calculate the div,maybe I just need use the second component of the grad rather than a whole vector . Are there any guides I can refer to? Then ,I have read the function of volumeInt().I find that it is total volume of a domain or subdomain rather than a cell volume.Are there any about the cell volume? regards lg88 

March 26, 2012, 04:51 

#10  
Super Moderator
Glenn Horrocks
Join Date: Mar 2009
Location: Sydney, Australia
Posts: 12,831
Rep Power: 100 
I have never used the .grad function, try it and you should be able to work it out.
Quote:
Cell volume is an available variable (see the CFX Reference manual) but again, I bet the volumeInt() function does what you want without requiring cell volumes. 

Tags 
divergence 
Thread Tools  
Display Modes  


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
Divergence problem for species transport model  MY  FLUENT  3  January 11, 2014 05:46 
Quarter Burner mesh with periosic condition  SamCanuck  FLUENT  2  August 31, 2011 11:34 
FVM in 1D spherical coordinates  lost.identity  Main CFD Forum  0  January 17, 2011 10:06 
comments on FDM, FEM, FVM, SM, SEM, DSEM, BEM  kenn  Main CFD Forum  2  July 18, 2004 18:28 
FVM: how to obtain the values on vertex?  remi Demersseman  Main CFD Forum  4  May 20, 1999 06:29 