CFD Online Discussion Forums - How does "solve" method works ?

Hello everybody, I have a question about the solve and relax methods.

1) It might be stupid but how does the solve method choose the "variable" to solve ?
For example in the compressibleTwoPhaseEulerFoam solver, if we take a look at the UEqns.h file, we can see the following code :

Code:

00029         U1Eqn =

00030         (

00031             (scalar(1) + Cvm*rho2*alpha2/rho1)*

00032             (

00033                 fvm::ddt(alpha1, U1)

00034               + fvm::div(alphaPhi1, U1)

00035               - fvm::Sp(fvc::ddt(alpha1) + fvc::div(alphaPhi1), U1)

00036             )

00037           - fvm::laplacian(alpha1*nuEff1, U1)

00038           + fvc::div(alpha1*Rc1)

00039          ==

00040           - fvm::Sp(dragCoeff/rho1, U1)

00041           - alpha1*alpha2/rho1*(liftForce - Cvm*rho2*DDtU2)

00042         );

00043 

00044         U1Eqn.relax();

How does it know that it has to solve U1 ? And how does it know that other variable must be "constant" ? Why not solving alpha1 ? Is it because U1 is the last argument ?

2) what is the difference between the "relax" method and the "solve" method ?

thank you in advance for your answer