There have been lot of discussion about CDS being unstable due to the reason that it can give unphysical oscilation in solution. I was wondering whether this is the reason for it:

when we descretise the equations we get (ap_conv + ap_diff) phi_p = sum(a_i * phi_neigh) + SU

now if (ap_conv + ap_diff) ~ 0 we get very high values of phi_p if we are solving by gauss seidel etc. that is we can not write the equation as : phi_p = ( sum(a_i * phi_neigh) + SU) / (ap_conv + ap_diff) as needed by Gauss Seidel.

Is this the reason for unstablity. Could someone comment on it.

I'm wondering: "raaxz" means "zxaar" ?

No i did not post it. But its good observation

U r correct its reverse of zxaar, because it came to my mind when i was reading some message by zxaar, so could not think of different name to put on, so just used reverse of it.

!noitanigami fo kcaL