CFD Online Discussion Forums

CFD Online Discussion Forums (
-   FLUENT (
-   -   Volume Flow Rate and Area * Velocity values different (

narendra.gadwal July 9, 2013 00:55

Volume Flow Rate and Area * Velocity values different

I have a fluent model, where the fluid is of constant density.
Structured mesh, well converged.

But the volume flow rate what fluent shows is different when I multiply the area and area weighted average of velocity magnitude.

Did not understand ? Why ?

(Volume Flow Rate) Fluent != (Area) fluent * (Area weighted average of velocity magnitude) fluent

Thanks in advance

RodriguezFatz July 10, 2013 03:41

Is it multiphase?

narendra.gadwal July 10, 2013 04:41

No .... the fluid is air

RodriguezFatz July 10, 2013 04:46

That sounds strange... How much do they differ?

RodriguezFatz July 10, 2013 04:49

Ahh maybe this:
What do you integrate for the "Area weighted average of velocity magnitude"? You need to integrate only the perpendicular (to the surface) component of the velocity vector - not the length of the complete velocity vector!

blackmask July 10, 2013 04:54

(Volume Flow Rate) Fluent == (Area) fluent * (Area weighted average of normal component of velocity) fluent

Note the difference between <vector u> dot <vector A> and |magnitude u|*|magnitude A|.

narendra.gadwal July 10, 2013 04:54

I tried with Velocity magnitude, X, Y and Z velocities .... still they differ.
When I calculate velocity = (volume flow rate/area) fluent and
area weighted avg of (velocity magnitude/X-vel/Y-vel/Z-vel), both differ by almost 4-5m/s

RodriguezFatz July 10, 2013 04:57

But this isn't what I wrote:
You need the perpendicular velocity component. If your surface is no "simple" cartesian surface, the cartesian (x,y,z) components won't help you.
What you need is "A dot v"!

narendra.gadwal July 10, 2013 05:10

It is a simple rectangular surface. So don't you think the velocity magnitude is normal to that ?

RodriguezFatz July 10, 2013 05:12

Not necessarily... Why don't you post pictures?

blackmask July 10, 2013 05:29

Let the normal of the rectangle be [nx, ny, nz], and the area-weighted average of x-,y-,z- velocity be [U, V, W], then (U*nx + V*ny + W*nz) should not be differ much from the volumetric flow rate divided by the area.

narendra.gadwal July 10, 2013 11:27

Thanks !
Will do that.

All times are GMT -4. The time now is 13:10.