Stokes plane-parallel vortex systems in channels (Q5931013)

This is an interesting paper reporting on new closed-form exact solutions of the biharmonic equation \(\Delta\Delta\Psi( x,y)=0 \). The function \(\Psi\) is interpreted as streamfunction of plane-parallel Stokes flows in two infinite domains: a) a channel with parallel walls (\(-\infty\leq x\leq\infty ,-1\leq y\leq 1\)), and b) a wedge-shaped region between two concurrent straight walls (\(0\leq r\leq\infty ,-\alpha\leq \phi\leq \alpha\)). Complex variable formulation and Goursat formula are used to obtain the solutions. The imposed boundary conditions comprise vanishing of the sought biharmonic function and its normal derivative on the walls. Streamline form or any condition at infinity are not specified in the course of the solution. In both cases, the produced types of solutions are vortex motions generated by the singularity (vorticity source) located at \(+\infty\). All solutions have physical significance only far away from infinity, because with increasing \(x\), respective \(r\), the assumption of slow flow does not hold any more. In the case a), an odd and an even solution for \(\Psi\) with respect to the \(y\)-axis are found. The well-known existence and uniqueness theorem concerning the solution of biharmonic equation in finitely bounded domains does not apply to infinite domains. Other known solutions to the Stokes problem for infinite domains considered here are Poiseuille channel flow and Hamel corner flow, which are obtained by assuming a priori a specific streamline form. On several graphically presented examples the authors also discuss the interaction between Poiseuille or Hamel flow and the respective vortex flows found in this work. Note that \(\Psi\) in the case considered in Section 1. is an odd function with respect to \(y\), not an even one as mistakenly translated. Its trigonometric part is periodic in \(x\) with the period \(2\pi/b\), not as it stands with the period \(2\pi/a\). The third part of Fig. 3 seems to refer to \(U=0.01\), not to \(U=-0.01\).