[m2montazari]

Hi,
I'm somewhat newbie in foam. I wanted to have a solver for unsteady coupled solver with variable density. I started with foam-extend-4.0 and simply edited pUCoupledFoam solver. At the first step, I tried to add time derivative. So added

 + fvm::ddt(U)

in UEqn.H and compile it to a new solver name, say "unsteadypUCoupledFOAM". But the result is an unstable solver which diverges for a simple starting poiseuille flow(which icoFoam doesn't have any trouble in convergence).
where is my fault? should I change any other things?
Also is there a better choice instead of pUCoupledFoam to start from?
thanks in advanced...

UPDATE:
I decided to check a more simple test case. So I ran first stockes problem (suddenly moving plate in semi-infinite domain). I overcame on the unstability by lowering timestep and using upwind scheme.
Now the problem is that icoFoam results are in good agreement with analytical solution; but unsteadypUcoupledFoam has some errors and do not fit on analytical solution.
I tried to extract rUA from convective-diffusive terms without time derivative term; but it almost doesn't change the results.