Flow coefficient according to IEC 60534 vs CFD simulation software

[fabiobs]

Hello every one. I have a problem with flow coefficient of ball and butterfly valve.
I tried to calculate flow coefficient of some butterfly valves according to IEC and simulating the scheme indicated. All values obtained with CFD are very close to real values but other competitors declare almost double values for same valve. I also tried to calculate kv value of ball valve (not far from a pipe with 8 nominal pipe diameters plus face to face length of valve).
Also in this case values obtained with CFD compared to values declared by ball valves manufacturer are very far.
In the simulation I have a 3D pipe with my valve, fluid is water ambient temperature, dP of 1 bar from inlet to outlet and software calculate the volume flow rate. Is there something wrong?

 

[SPDL310]

Fabiobs,

I have seen similar competitor claims in for choke valve Cv values. In the one specific instance I am thinking of it turned out to be marketing claims for a vapor ware valves that was still under development. It may be the same in your case, but there is always a non zero possibility that you analysis is incorrect. A few questions to ask yourself:

1.) Is there one specific competitor making the claim, or is it every body else.

2.) Has you company ever physically tested a valve that you can use to validate you CFD methodology?

3.) How do the Cv values you generated stack up the the Cv values of an orifice diameter of the equivalent flow area exposed?

You may want to make a more detailed post in the CFD forums to see if it is something on your end. Much like the physical test methods for Cv I would change the flow parameters you are looking at to determine your Cv. If you are only going for a 1 bar dp across the valve you may be in a laminar flow regime. I would run a range of dp's and flow rates to see if you are getting consistent results, and use the average of those values to determine a Cv. At the end of the day you need to validate your CFD method by comparing a physical test to simulation results in order to have confidence in your simulation.

 

[fabiobs]

Thank you for your reply.
I think that all competitors try to declare a higher value than the real one. We had a test bench used to test small valves (up to 1"1/2) and CVs obtained in the test are +/- 4% CVs obtained in CFD analysis.
The problem is that competitors values are nearly twice as much.