[cmbv]

Hi
I am trying to simplify a multi-component, high temperature and pressurized flow into air at atmospheric conditions.
I heard and have read that two flows are similar as long as their Reynolds numbers match.
So I set up a very simple model –a single 25cm channel with a cross sectional area of ~1x1 mm.
I ran this model for the flow with all the components, at a high pressure and a high temperature and then I also ran it for air at atmospheric conditions.
Then I did this very simple calc:
m2 = (mu2*m1) / mu1 –where m’s are flow rates and mu’s are viscosity 1 & 2 refer to the two different flows (gas mixture and air)
This are not remotely turbulent they are both in fact very laminar, and both Reynolds numbers match very well.
However when it comes to measuring the pressure drop, they defer by almost an order of magnitude: 7 Pa (air case) and ~70 Pa.
Why is this? Is it not possible to use this for measuring pressure drops this way? How about the distribution of flows, in case there are splits and separations; will they both give you similar right results?
I have tried all sorts of things, most diffusion models, running all the turbulent models turning turbulence off, running incompressible flows, inviscid, double precision, etc… All of these have given me an order of magnitude difference.
Do you know why?
Regards
cmbv