Grid Convergence

As far as I know, the influential factors in a physical model are its dimensionless numbers such as Re,Pr,Ra,Ri,…
But, in my code whenever I change the inlet velocity (ulb , velocity in lattice unit) while maintaining the dimensionless numbers (Re=100, Pr=0.71, Ri=1,) constant , the results vary. For example, if the grid size changes (ulb=Redxnu) ulb also will change so the results will be different . This situation is also the same when the grid size is constant and I, myself change the velocity inlet.
Somebody please explain to me why it is like this or what am I doing wrong?
Tnx.

Probably you don’t do proper dimensionalization or your non-dimensional parameters (which you keep constant) are non-leading physical parameters, or your grid parameters are too small.

Please check the governing non-dimensional parameters for your problem or post more details,
Cheers,
Alex