Can circum-avg-axial give flow rate?
 

I've been using Fluent for a while now, and I've noticed that something is very odd about the circum-avg-axial function. In simulating incompressible flow through any internal geometry, the average value of axial velocity should remain constant (with a weighting factor of the local radius). In simulating a pipe constriction, resulting in backflow, the average of axial velocity should remain also remain constant with weighting of r (integral of u dot n r dr d theta= flow rate = constant).

When I run circum-avg-axial on the constriction geometry ( axial velocity times the radius) however, I do not get a constant value. The value fluctuates wildly, especially around the region of the geometry change.

I have tried the circum-avg-axial of: axial-velocity axial-velocity * radial coordinate axial-velocity * radial coordinate ^ 2

I do not think any of these is giving me the equivalent to the local flow rate.

Why does circum-avg-axial give this wrong output? Or what function combination should I be using to calculate the flow rate from axial velocity at a given axial distance down my pipe?

It is important that I get flow rate at a specific axial distance-- such as z = 1 or 2, and not flow rate on the entire interior.

Thanks in advance for your help, Amit