I ran the script and I noticed the following issues. First of all "vel" is
not a vector and is required to be in order to pass it to a convection term.

The shape of "vel" is (161,) and is thus a scalar. Now the operation

     discharge=dot(head.getFaceGrad(),kond.getHarmonicFaceValue())

retunrs a scalar (161,). What are you trying to do with that operation?
kond.getHarmonicFaceValue() is a scalar and head.getFaceGrad()
is a vector so I am not sure what a dot operation means there.

When I make the following changes things seem to run. Of course I have no idea
if the result is what you are look looking for.

46c46,48
< discharge=dot(head.getFaceGrad(),kond.getHarmonicFaceValue())
---
> ##discharge=dot(head.getFaceGrad(),kond.getHarmonicFaceValue())
> discharge=head.getFaceGrad() * kond.getHarmonicFaceValue()
>
47a50
>
60,62c63,65
< transportEq2=(TransientTerm()==
<               -PowerLawConvectionTerm(coeff=vel())
<               -DiffusionTerm(coeff=Disp))
---
> ##transportEq2=(TransientTerm()==
> ##              -PowerLawConvectionTerm(coeff=vel())
> ##              -DiffusionTerm(coeff=Disp))
63a67,69
> transportEq2=(TransientTerm()==
>               -PowerLawConvectionTerm(coeff=vel)
>               +DiffusionTerm(coeff=Disp))


On Jan 24, 2008 11:17 AM, asreeve <[EMAIL PROTECTED]> wrote:
> Daniel,
>
> I've attached the file. Its a very simple mesh...just one cell in each
> layer, so I could easily keep track of what's going on and check it
> against a 1-D analytic solution. I've attached both the mesh and geometry
> files.
>
> Andy
>
> ----------------------
> Andrew Reeve
> Dept. of Earth Science
> University of Maine
> Orono, ME 04469
>
>
> On Wed, 23 Jan 2008, Daniel Wheeler wrote:
>
> > Date: Wed, 23 Jan 2008 16:59:23 -0500
> > From: Daniel Wheeler <[EMAIL PROTECTED]>
> > Reply-To: [email protected]
> > To: Multiple recipients of list <[email protected]>
> > Subject: Re: using convection term
>
> >
> >
> > Hi Andrew,
> >
> > Can you attach './HalfCirc.msh' so I can run the script.
> >
> > Cheers
> >



-- 
Daniel Wheeler

Reply via email to