[Nek5000-users] Boundary condition for a passive scalar?

[nek5000-users at lists.mcs.anl.gov]

Hello,

I am trying simulate a backward facing step flow at Re = 5000 with a
passive scalar using Schmidt number = 0.71. I am having some trouble with
the boundary condition for the passive scalars.

At the wall I use zero gradient boundary condition by using insulated
boundary ("I  ") as the flag in the .rea file. After several flow thorough
times, the passive scalars just fill the domain up and become uniform
everywhere. This can be observed from the following 2 movies.

1. This is the contour plot of the passive scalar. The "Min" value of the
scalar concentration keeps increasing.
http://www2.uic.edu/~hkanch1/s2.mpeg <http://www2.uic.edu/~hkanch1/s3.mpeg>

2. In this movie, the "Min" value of scalar is fixed. It can be observed
that the scalar concentration is becoming uniform throughout the domain.
http://www2.uic.edu/~hkanch1/s3.mpeg

The simulation works just fine if I set the passive scalar value to "zero"
on the wall.
http://www2.uic.edu/~hkanch1/s4.mpeg

I don't understand why this is happening. May be something wrong in my set
up of the problem.

Regards,

Harish.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/nek5000-users/attachments/20120207/3fe7feb8/attachment.html>