Zou-He or bounce back?!

Hi Dears
I have confused a n important question: for walls of channel boundary condition Can we implement Zou-He with(u=v=0) instead of bounce back : (f(i)=f(opposite(i))??

6 2 5
\ | /
3----0----1
/ |
7 4 8
for example for upper wall Zou-He BC yields:
f(2)=f(4)
f(5)=f(7)-0.5*(f(1)-f(3))
f(6)=f(8)+0.5*(f(1)-f(3))

–buonce back yields:
f(2)=f(4)
f(5)=f(7)
f(6)=f(8)

I tried two above methods for a uniq problem. the results are different (with the first BC, flow reachs fully developed in shorter length)and it seems that the first method is more accurate. what is your opponion??
What about half bounce back? I don’t know how to programmed half bounce back. Does anyone have a part of code showing how to imlement half bounce back??

Hello,

The Zou-He wall is exactly located on the wall node, but BB wall is not. The BB wall is located is somewhere between BB walls and fluid nodes (this depends on the relaxation parameter) giving the second order accuracy for the location in the middle. If you are interested to obtain the exact proper wall location you need to use either Zou-He or other more precise wall conditions or use BB wall with the certain relaxation parameters for certain flows.

You can read about implementation here - http://www.lbmethod.org/forum/read.php?3,2375,2535

Hopefully, it will help,
Alex