Stability and convergence of some parallel iterative subgrid stabilized algorithms for the steady Navier-Stokes equations (Q2673505)

The authors claim that using the finite element method and a fully overlapping domain decomposition method they ``propose and study some parallel iterative subgrid stabilized algorithms for the simulation of the steady Navier-Stokes equations with high Reynolds numbers''. The first observed inadvertence is that neither the Navier-Stokes system, in its original form, nor its weak formulation do not contain the Reynolds number but only the kinematic viscosity. Moreover, I do not think that the case $f:=0$ can be treated with the theory presented by the authors, but the last two numerical examples are given exactly for this case. The Theorem 2.1 and the Lemmas 2.2, 2.3, etc. derived from it, are not valid in this case. It is not at all clear what the constant $N$ is. The authors simply say: ``a constant related to the convective term depending only on $\Omega$''. An analogous discussion can be made for Lemma 4.3. The strong condition of uniqueness is trivially fulfilled for $f: = 0$, but the conclusions of the lemma no longer take place. In addition, we do not see how this strong condition of uniqueness can be satisfied for large Reynolds numbers. We believe that the conclusions of this paper must be taken with great caution.