Developing Flow Friction Factor

[arkie87]

All,

I have posted this same question in a cfd-online forums. But i figured i'd repost here, since the question is pretty basic, and there should be a lot of external flow people here (and developing internal and external flow are pretty similar for really short channels).

Anyway,

I am trying to simulate developing flow between two parallel plates (2-D simulation), and compute the dimensionless developing (Darcy) friction factor Reynolds product over fully-developed value of 96 (f*Re/96) as a function of dimensionless length down the channel (X^ = X/(Re D)).

From the literature, it seems that the curve of f*Re/96 vs X/(Re D) should be constant and independent of Re; however, in my simulations, in multiple softwares (ANSYS Fluent, COMSOL, my own Matlab code), i find that the curves vary with Re, and increasing Re decreases f*Re/96 for the same X/(Re D).

I compute f two ways, both of which give nearly identical results:

(1) dP/dx = 1/2 f/D rho V^2; solve for f

(2) f = 4*mu*du/dy|w / (rho V^2) (factor of 4 to convert from Fanning to Darcy)

p is static pressure, but ive tried total pressure as well, and it doesnt fix the problem.

Is this correct? What am i doing wrong?

Would appreciate any help,