[mortezanayebi]

Originally Posted by gschaider

The usual behaviour of the internalField-parser (which is used by the expressionField-function object) is to only calculate the internal field (hence the name) and calculate a zeroGradient for the patches (which basically means that a face gets the value of the neighbour cell). Why that is 0 in your case ... I don't know (unless nearDist is 0 in the boundary cells)

Anyway: variation 1 is better anyway.
a) because it only has to be computed on the boundary. Not on the whole field
b) you get yPlus for the face. Not the value of the cell projected onto the face. Which might differ (think of a stretched cell in a corner: all three boundary faces of that cell will get the same value. Although they don't have the same yPlus)

There is another way to calculate that. You could use the readAndUpdateField-functionObject to load a field myYPlus. In that field you implement the expression from version 1 as a groovyBC. Then you can use the patchExpression from version 2. The only advantage of this is that afterwards you have the whole field on disc and can visualize the distribution of yPlus in paraView

Hi Bernhard
I need to find local nusselt around a cylinder. I was wondering if you could please let me know how I can obtain the local nusselt using swak4Foam.

I might need to mention that I already calculated the average nusselt by adding the following code to controlDict.

libs (
      "libOpenFOAM.so"
      "libsimpleFunctionObjects.so"
      "libsimpleSwakFunctionObjects.so"
      "libswakFunctionObjects.so"
     );
functions
{
             NusseltNumber
            {
                        type patchExpression;
                        ignore_unimplemented_simpleFunctionObject::movePoints    true;
                        outputControlMode   timeStep;
                        outputInterval    1;
                        patches
                        (
                                    cylinder
                        );
                        verbose true;
                        variables
                        (
                                    "T_inf=0;"
                                    "L=10;"
                        );
                        expression "L*snGrad(T)/(T-T_inf)";
                        accumulations
                        (
                                    average
                        );
                        log         true;
            }
}

thanks

[wyldckat]

Quick answer: Looks like the only way to do this is to use groovyBC to do the calculation on a field. I don't remember ever doing this myself and right now I don't have enough time to come up with an example case.

That said, it might be easier if you modify the old source code applications for wallHeatFlux and modify it to your needs...

Wait, I did a bit more searching and there is already a complete example here: calculate Nusselt number - post #2