[ooo]

I've added some lines of codes in icoFoam, After solving the momentum equation (and before of PISO(pressure correction) loop) .
I use the 'U' field(which has already solved in momentum equation) and do some modification on that to get a new 'U'. now, this new 'U' is given to the piso loop.
The confusion part for me is that:
In some parts of PISO loop, the .H and .A coefficients of the UEqn are used.
In my code, i've eddited the 'U' field before the PISO loop, so is this a correct way to again write the UEqn with respect to the new 'U' ? I've done it but the result is not changing and i wonder if i miss something after computing the new 'U' and before the PISO loop :

...
1)fvVectorMatrix UEqn(fvm::ddt(U)+ fvm::div(phi, U)- fvm::laplacian(nu, U));
2)solve(UEqn == -fvc::grad(p));
3)Using the 'U' from '2' and do some modification on that
4)fvVectorMatrix UEqn(fvm::ddt(U)+ fvm::div(phi, U)- fvm::laplacian(nu, U));
5)Piso loop 
...