Hello,

I'm working on flow calculations with RANS 2eq models and I'd just like ask : the term d/dx(-2/3*densit*k) is dropped from the momentum eq's quite often in the literature without any comments? Why? Is this generally an accepted approximation or is it because people want to make their simulations more stable? Some codes have included this term into the pressure gradient (giving a modified pressure) but I guess creates a special treatment when pressure is used elsewhere. Any hints on this anybody?

Regards

Andreas

As far as I know, it's almost standard to treat this term together with the pressure, and some people (including myself) don't even bother taking it out afterwards (small relative to the pressure level). The thing is that for stability (at least in collocated FV codes) it requires exactly the same treatment as the pressure gradient and it is easiest just to add the terms together. I have never seen this cause any trouble; if anyone has, I'd be interested.

Hrv

Thanks, Hrvoje! Is the incorporation of this term into the pressure also "standard procedure" when using nonlinear eddy viscosity or explicit ASM models?

/Andreas

Hi,

Yup, I use it always, even for Reynolds stress. It has never caused me trouble in non-linear k-e, but i'm not sure whether they are really worth the effort.

Have fun,

Hrv