[Venu Angirekula]

Hello everyone,


This question is basic and theoretical and independent of system and software used.

For explanation, I have taken well known elbow tutorial and considered two scenarios with two different inlet velocities. For Case1: U1 = (2, 0, 0) and for Case2: U2 = (2, 2.5, 0). After completing the simulations, I got the expected flow behavior, but when flowrate(phi) is evaluated at the inlet, I got same values for the both cases. Please check the attached image.

flowrate = velocity * Area
Area = 30.052 m2;

For Case1: Q1 = U1 * Area = 60.153 m3/sec
For Case2: Q1 = U2 * Area = 60.153 m3/sec

I understand from here, that for calculation of the flowrate, only the velocity component to inlet patch is used here, which is x-component.

Finally, the question is,

How the remaining Inlet velocity components are going to be used for faces fluxes with in the system?.

[FMDenaro]

I have doubts about the sense of your question. The flow rate is always a scalar quantity that depends on the component of the mass flux normal to the area. On the other hand the mass flux is a vector.

[PenPencil]

I have doubts about the sense of your question. The flow rate is always a scalar quantity that depends on the component of the mass flux normal to the area. On the other hand the mass flux is a vector.

This is not true. You can formulate both the volume flux and the mass flow rate using vectors. This is as simple as quantities that have direction and modulus.


Venu, I believe the other components are not going to matter to the inlet face flux, as they do not contribute to it and as you have noticed.
However, it is not because the other components have no influence on the face flux that they will not disturb your flow. They certainly will, because at the end of the day you have a diagonal flow entering your domain rather than a horizontal one if you have a non-null velocity-component.

[FMDenaro]

Originally Posted by PenPencil

This is not true. You can formulate both the volume flux and the mass flow rate using vectors. This is as simple as quantities that have direction and modulus.


Venu, I believe the other components are not going to matter to the inlet face flux, as they do not contribute to it and as you have noticed.
However, it is not because the other components have no influence on the face flux that they will not disturb your flow. They certainly will, because at the end of the day you have a diagonal flow entering your domain rather than a horizontal one if you have a non-null velocity-component.

I would be glad if you state in a more rigorous way what you are doing.
I assume the mass flux to be rho*v, that is a vector field. The flow rate is the projection of the mass flux along the normal unit vector to the surface and integrated over the surface, that is the scalar quantity



Int[S] n.v*rho dS


while the volumetric flow rate is



Int[S] n.v dS




In conclusion, you need to have a curved surface (that is n=[nx,ny,nz]) to get the contribution of all velocity components.