CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (http://www.cfd-online.com/Forums/openfoam-solving/)
-   -   TimeVaryingUniformFixedValue boundary condition (http://www.cfd-online.com/Forums/openfoam-solving/59387-timevaryinguniformfixedvalue-boundary-condition.html)

liu May 9, 2007 15:01

Correct me if I am wrong. But
 
Correct me if I am wrong. But the timeVaryingUniformFixedValue for vector (such as veloctiy) on a boundary can only specify the magnitude. The direction of the uniform vector is acturally the surface normal:

template<>
void timeVaryingUniformFixedValueFvPatchField<vector>:: updateCoeffs()
{
if (this->updated())
{
return;
}

checkTable();

operator==
(
-interpolateXY
(
this->db().time().value(),
timeDataPtr_().x(),
timeDataPtr_().y()
)*patch().nf()
);

fixedValueFvPatchField<vector>::updateCoeffs();
}

All these boils down to the design of graph class. It can only accept x-y relationship where x and y are both scalars. Can't y be vector? That way we can specify a graph using data file like
(
t0 (v1 v2 v3)
t1 (. . .)
...
tN (v1N, v2N, v3N)
)

Interpolation of vector between t(i) and t(i+1) can be component wise. If graph is designed like this, timeVaryingUniformFixedValueFvPatchField<vector> can specify uniform velocity in any direction.

valentine October 12, 2007 14:19

timeVaryingUniformFixedValue b
 
timeVaryingUniformFixedValue boundary condition does not seem to work with sonicLiquidFoam. Any idea why this is the case?

Thanks.


All times are GMT -4. The time now is 02:54.