[waynezw0618]

hi every one�
i have some question about MRFZones.C
in void Foam::MRFZone::addCoriolis(fvVectorMatrix& UEqn) const:

const labelList& cells = mesh_.cellZones()[cellZoneID_];
const scalarField& V = mesh_.V();Q1:what is the mesh_.V()?
....

forAll(cells, i)
\ {
Usource[cells[i]] -= V[cells[i]]*(Omega ^ U[cells[i]]);
Q2:here coriolis force is equal to "omega × u",but i think it should be "1/2 * omega × u" and also there is no centrifugal force which is "omega×(omega×r)",so if i want to calculate the flow in the turbins? would i need to modify the them both?
}

thanks
wayne

[dmoroian]

Hi Wayne,
I think that in an absolute velocity formulation, the Coriolis force (2rwxU_relative) and centripet force (rwxwxr) can be written as: rwxU

I hope this is helpful,
Dragos

[dmoroian]

...and mesh_.V() retuns the cell volume

[waynezw0618]

Hi Dragos：
i don` think so.if we write the moment equ in rotating frame, the velocity is relative velocity (W),and in the absolute frame the equation is without the coriolis force and centrifugal force.
in MRF:

rho*D(W_relative)/Dt=F(total,including pressure,viscosity)+omega×(omega×r)-2* omega × u


in absolute frame:
rho*D(U_absolute)/Dt =F

according to you should be

rho*D(U_aboslute)/dt= F+ * omega × U

i dont think it will be
rho*D(W_aboslute)/dt= F+ * omega × U

for the solver just treat one velocity not two

thanks alot
yours wayne

[waynezw0618]

i am sorry, there is a "*" before "omega × U" for last two equ.it is typing error of mine

thanks
wayne

[dmoroian]

I had a vague memory in the back of my mind when I wrote the above post, but I've just dug for it and here it is...

In the absolute velocity formulation and for a steadily rotating frame the flow equations can be written as follows:

- continuity:

- momentum:

- energy:

where the relative velocity is:

and the transport velocity is:

You can also check the fluent user manual for it.

I hope this is helpful,
Dragos

[waynezw0618]

Hi Dragos
thanks a lot!
i could understand it now�
the relative velocity is as calculating in the memeber function relativeflux,so moment eqution should be div(phi,u),is that right?

thanks
waye

[dmoroian]

Hi Wayne,
This is exactly what it is written in the momentum predictor:

tmp<fvvectormatrix> UEqn
(
<blockquote>fvm::div(phi, U)
+ turbulence->divR(U)

);
</blockquote>

Dragos

[waynezw0618]

Hi Dragos
but why the flow pattern of the result looks like U is relative velocity

wayne

[dmoroian]

well Wayne, if it is like that I agree with you: it is strange.
Can you post a picture with an example?

Dragos

[waynezw0618]

Hi dragos
the result of mixvessl2D in tut looks likes the U is absolute velocity

[waynezw0618]

[waynezw0618]

[waynezw0618]

[waynezw0618]

Hi sorry
i try a lot
here is the result of mixvessl2D(vector on the rotor),you can see clearly the velocity profile relative to the omega

and then is my case a 3d centrifugal impeller
a vector plot of U in the impeller z=1/2thick

b vector plot on the blade

[waynezw0618]

mix2d

[dmoroian]

Lots of tries!
Well in the picture I can see, the vectors look ok to me!
Be careful to have pictures with dimensions of less than 50kb.

Dragos

[waynezw0618]

a vector plot of U in the impeller z=1/2thick

[waynezw0618]

b vector plot on the blade

[waynezw0618]

c relative velocity (at the same conditon) calculated by CFX