Modelled Part of Reynoldstresses in Channelflow
 

Dear Openfoamers,

at the moment i'm running a channelflow case with a very fine constant resolution of y+=1 ortoghonal to the wall and rough resoluton in stream and spanwidth direction of (around x+= 50 and z+= 20).
For this case the modelled part of the equation become very small, what for me seems not to be really astonishing, since the y direction should have the most influence on the modeled part, but my prof says it should be the other way arround and x and z should have the most influence and hence i should a really high modelled part. For example if i'm running a dynamic Smagorinsky modell with plane avaraging in OpenFoam, what would you say has the greatest influence on the modelled SGS part - the y-direction, the x-direction or the z-direction? Would you say with (very) small celllength in y-direction and large in every other direction the modelled Part should become large?

Kind regards

Le Frog