[nikoscham]

I use the chtMultiRegionFoam (openFoam v2012) to solve heat transfer between fluid and solid regions.
I have frozen the flow and at one boundary of the solid region (which is NOT in contact with the fluid), I impose a fixedGradient BC for the Temperature as follows:

    
Geom4_2f
    {
        type            fixedGradient;
        gradient        uniform 20;
    }

I expect that the above BC would result to a heat flux of:
q = gradT/Ksolid = 20/117 = 0.17 W/m2
where Ksolid = 117 W/(mK) and gradT = 20 K/m (according the to the fixedGradient BC)

Surprisingly enough, when I use the wallHeatFlux tool I get (for the same boundary) 2340 W/m2 ! The later can be also verified by the (large) increase of the temperature at the results.

What is going on there? Are the fixedGradient units wrong? Any ideas?