[JayDeeUU]

I'm creating a slightly modified solver based on buoyantBoussinesqPimpleFoam. What I want to do is modify TEqn.H by adding a term that (implicitly) represents vertical advection of a background temperature gradient, say, - w dT0/dz. The idea is to look at the effect of stratification on a weak, buoyancy-driven flow. I have been guided by, and tried to follow along with, these helpful forum entries:

https://openfoamwiki.net/index.php/HowTo_Add_a_Coriolis_Force_And_Make_a_Tornado

http://https://www.cfd-online.com/Fo...-equation.html


In transportProperties I define the background temperature gradient, e.g.

gradT0 [0 -1 0 1 0 0 0] (0.0 0.0 0.1)

and in readTransportProperties I added an entry

dimensionedVector gradT0
(
    "gradT0",
    dimensionSet(0,-1,0,1,0,0,0),
    laminarTransport
);

The modified solver compiles just fine, if I don't add my new term in TEqn.H. However so soon as I add the new term, viz. the last term shown below

fvScalarMatrix TEqn
    (
        fvm::ddt(T)
      + fvm::div(phi, T)
      - fvm::laplacian(alphaEff, T)
     ==
        radiation->ST(rhoCpRef, T)
      + fvOptions(T)      
      - U & gradT0
   );

the solver no longer compiles, and wmake produces many messages, the most easily comprehensible of which is:

TEqn.H:28:9: note:   mismatched types ‘Foam::DimensionedField<Type, Foam::volMesh>’ and ‘Foam::fvMatrix<double>’
       - U & gradT0

I understand that my gradT0, unlike U, is not a time-varying vector field, and that this may be related to the problem. I'm not an experienced C++ programmer so this is liable to be an elementary mistake on my part. Any suggestions as to how to fix the problem will be greatly appreciated. Thanks, John. (Note: I'm using OpenFOAM 2006.)