Hi all, I want to model H2 flow through a microchannel. I am feeding the channel at 25 psi and exit is at atmospheric. There are several bends. I would like to model this flow. Should I use the Euler equation or the compressible Navier Stokes? Which one gives a better predictions? i calculated the velocity using an experiment measuing the volumetric flow rate, the average velocity seems to be 60 cm/s, which is well below the Ma number of 1320 m/s in hydrogen. thanks,

Adarsh.

Maybe you can consider the DSMC method and you can get code from Bird`s book.

I don't know how small (or large) a microchannel is.

Do you have a continuum flow? Is the mean free path a sizable fraction of the diameter (width, height)?

The Knudsen number, K = mfp/dia, should be small (< 0.1) to use the continuum (Navier-Stokes) model.

This stuff is coming off the top of my head, and I'm sure folks with more experience will correct or refine as needed.

If your knudsen number is less than 10^-3 use a continuum model and if the knudsen number is in the range 10^-3 - 0.1 use a continuum model with slip boundary conditions

u_s=u_wall+du/dn)wall If Kn> 0.1 you can still consider slip flow model upto a Kn of 4-5 and beyond that use DSMC as mentioned by dusky.

-H