fvMatrix.A()

santiagomarquezd · December 20, 2010, 16:03

Hi all, I was wondering why .A() divides the diag() values by mesh.V() values. Another question is how only n values are obtained by A even from a fvVectorMatrix which has 3*n equations for 3*n unknowns.

Thanks in advance.

santiagomarquezd · December 25, 2010, 00:17

Regarding the first question I was doing some dimensional analysis and it shows that in order to do things like:

Code:

00096             U -= rUA*fvc::grad(p);

in icoFoam.C it is necessary to remember that matrix coefficients in FVM are obtained from volume integrals, then if one operates with no integrated quantities like fvc::grad(p) this coefficients have to be divided one by one by the corresponding cell volume.

Can anybody confirm that?

Regards.

sharonyue · June 6, 2013, 23:16

Hi Santiago, I think my problem is the same with yours, http://www.cfd-online.com/Forums/ope...n-icofoam.html Did you find the solution?

Thanks

December 20, 2010, 16:03	fvMatrix.A()	#1
santiagomarquezd Senior Member Santiago Marquez Damian Join Date: Aug 2009 Location: Santa Fe, Santa Fe, Argentina Posts: 452 Rep Power: 23	Hi all, I was wondering why .A() divides the diag() values by mesh.V() values. Another question is how only n values are obtained by A even from a fvVectorMatrix which has 3n equations for 3n unknowns. Thanks in advance. fumiya likes this. __________________ Santiago MÁRQUEZ DAMIÁN, Ph.D. Research Scientist Research Center for Computational Methods (CIMEC) - CONICET/UNL Tel: 54-342-4511594 Int. 7032 Colectora Ruta Nac. 168 / Paraje El Pozo (3000) Santa Fe - Argentina. http://www.cimec.org.ar

December 25, 2010, 00:17		#2
santiagomarquezd Senior Member Santiago Marquez Damian Join Date: Aug 2009 Location: Santa Fe, Santa Fe, Argentina Posts: 452 Rep Power: 23	Regarding the first question I was doing some dimensional analysis and it shows that in order to do things like: Code: 00096 U -= rUA*fvc::grad(p); in icoFoam.C it is necessary to remember that matrix coefficients in FVM are obtained from volume integrals, then if one operates with no integrated quantities like fvc::grad(p) this coefficients have to be divided one by one by the corresponding cell volume. Can anybody confirm that? Regards. __________________ Santiago MÁRQUEZ DAMIÁN, Ph.D. Research Scientist Research Center for Computational Methods (CIMEC) - CONICET/UNL Tel: 54-342-4511594 Int. 7032 Colectora Ruta Nac. 168 / Paraje El Pozo (3000) Santa Fe - Argentina. http://www.cimec.org.ar

June 6, 2013, 23:16		#3
sharonyue Senior Member Dongyue Li Join Date: Jun 2012 Location: Beijing, China Posts: 838 Rep Power: 17	Hi Santiago, I think my problem is the same with yours, http://www.cfd-online.com/Forums/ope...n-icofoam.html Did you find the solution? Thanks