HI!

I simulate a microgap flow (height 100um). With k epsilon turbulence model. First my mesh's height consists of 10 layers totaly. my values for y+ have been 7<y+<20 By now I refined my mesh (16 layers totaly over height) and y+ was reduced to 3<y+<11. This is logical, regarding that the first node's distance from the wall was reduced. But the modeling guide says that I have to avoid y+ values smaller than 20. I guess this has to be because the logarithmic wall law needs a minimum space to be valid. How Can I refine my mash without getting in trouble with too low values of y+?

Thanks for some ideas or help

Seve

If you wanna to use k epsilon model, the y+ must be much bigger than 10, because k epsilon model use wall function at near wall region. You can find more instructions in the CFX help. After all, what's your problem?

i think, for such a narrow gap, is there any necessity to use turbulence flow model?

Is using a different turbulence model not possible that desires lower Y+?

Microgap flow must be laminar!

SST can resolve boundary layers if Y+ less than 2. So you need to make the mesh finer.

But then if microgap flow is laminar, things would be different , wouldn't they?

y+ is a turbulent parameter. It doesn't exist in laminar flow.

I am not sure if this question has been asked, but I have always wonder...

If you have enabled any one of the turbulence (say, just the RANS) models, but most of your flow parameters are in laminar condition. How does the solver actually treat it? Would that give un-physical results?

From Omer's question, it led me to think of the laminar sublayer that exists near the wall under turbulence flow condition. So, when y+ < 2, the flow in that region is laminar, does that same turbulent model is being applied there?

thanks in advance for answering.

jak