[C.C]

Hi all,

I want simulate a turbulent pipe flow of a non-Newtonian fluid... I'm looking for the representation of U+ as a function of y+, because I'm going change the log wall law taking into account a reference in the literature.
The non-newtonian behavior was translated by a power law (viscosity as a function of shear rate) that is available in Fluent for turbulent flow...
In the figure, the green line corresponds to the standard log-law considering the viscosity constant.
The U+ presents a strange behavior... Someone knows why I obtained this behavior, some references about the non-newtonian fluids and turbulence in CFD or how the non-newtonian behavior is implemented in the k-e turbulence model? Thank you

[C.C]

I've calculatd y+ with density and viscosity in each node, but I found a paper where the authors considered the mean wall viscosity instead of Newtonian viscosity. If I consider the mean wall viscosity the profiles seem be OK... what strategy is correct?

Someone knows if in the dimensionless distance (Attached Expression) the density and the viscosity when the fluid is non-newtonian is evaluated in each node or can be equal to mean wall viscosity?