How can the velocity at points of change in pipe systems be determined?

[pipedown]

For minor losses, I read that the pressure drop is equal to the dynamic pressure * loss coefficient. However, in calculating dynamic pressure, it uses a single velocity. How do you determine this velocity when it will vary at the point of change in pipe diameter or direction? I have read somewhere that it would be a mean velocity between the two sides of the change, but how can this be determined? It seems that in these pipe systems, the head loss must be calculated for the system to determine velocities before a mean velocity at these points can be calculated. Is an iterative process the only way to achieve this or am I missing something?

 

[Skrambles]

If you know the volumetric flow rate you can just divide it by the cross-sectional area of the pipe. It really depends on what information you have about the system, sometimes you will have to find the answer through iteration and sometimes you may need to use a table.

 

[Curl]

Actually this depends on how the coefficient is defined. Usually the table you grab it from will tell you if they are referring to the velocity coming out, or the upstream velocity.

 

[pipedown]

Is there a general convention? I am also trying to figure out how to relate this in terms of reservoirs/tanks and entrance flows, though I am starting to see that I should be making assumptions for certain head losses based on relative geometries