Hi
How to use 'rho' in a scalar transport equation. If i use like following,
ddt(rho, T)+div(phi,T)+laplacian(K,T), I have dimension problem..
T=dimLess
ddt(rho, T)= 1/s kg/m3

phi=linearInterpolate(rho*U) & mesh.Sf()
div(phi,T)=1/m kg/m3 m/s m2
=kg/m-s
Therefore different dimensions. In rhoTurboFoam momenturm equation is working with this definitaion. I could not understand where is my mistake. Anyone help me please !!

Regards
-Liton

[stephan]

hi,

i think if you using the same argument for the momentum equation in rhoTurbFoam you will find the same problem.
it's the m^2-term which is not there: writing the transport equations with div for the convection implies that this term is an volume integral
(gauss:int (dui/dxi ) dV =int ui dAi)
i guess the problem with the dimension comes from the laplacian term.
what dimension has k in your set up (there should be no kelvin in the dimension, [k]=kg/(s*m)) ?
stephan

Hi Stephan Gerber, Thanks for your reply. actually i wrote the equation in a wrong fashion. When I printed the phi, I found its dim is 1/s .. as mesh.Sf() is area vector I thought the dimension of area must be multiplied with the U.

one more question. if I declare a volScalarField like following:

volScalarField XXX
(
IOObject(
..
..
),
c1+c2
)

where c1 and c2 dimensionedScalar. If c1 or c2 update, XXX won't update automatically. How to get the update of XXX. I used like following
XXX=c1+c2;
It works, but it does not seem to me very convenient. Any tips!!

Regards - Liton

[stephan]

hi,

you may combine calculating XXX=c1+c2 with calling XXX (some kind of special-output routine for XXX) which calculates the equation - so you wont have to calculate explicitly.
maybe there is a nice way using pointers or references?
stephan