[Duarte Albuquerque]

in Jasak PhD http://www.h.jasak.dial.pipex.com/papers.html
he uses different analytic solutions to test the refinement procedure.

in the line jet problem page 212 he uses 3 inlet boundary conditions and 1 outlet boundary condition in the rectangle domain.

Typically you can use zero gradient in the outlet boundary for moment condition. But if I derived the analytical solution of gradient in the outlet boundary the value is not zero it varies with y. This affects the error distribution close to boundary, which is different from Jasak plots.

Should I change outlet for linear extrapolation
of velocity does that work?

Someone has a clue how can i solve this problem?