[SteveM]

Hi Guys,

I am new here, great to see a community for things like this.

I have found this place through coming up against an issue I seem unable to solve.

I have a pipe line assembly that we have physically tested, we introduced the flow to the pipe and monitored this along with the pressure drops, temps etc.

The real results showed that at 295.57 US Gal / min there was a pressure drop of 1 PSI naturally occurring, so after calculation would result in a CV value of 295.57 aswell.

Now, I have taken all the data, worked up an average and also ran individual tests to work out the compensation for losses in the test equipment and get an average CV value of 304. So not too far away.

I have simulated this in floefd as an external flow with the inlet capped and set at X US Gal/min, the pipeline follows with the same pressure point indicators as the test equipment, with an extended tube after that running out to atmosphere. (This was the only way I was aware of how to input just the x US Gal/min and get a result) anyway, these results came out to within 1% of the real tests, so at that point I am quite comfortable that the CV would be 295.57 or there abouts.

HOWEVER,

When I am running a CV check on other assemblies which I have not tested, my normal set up is to have inlet and outlet tubes running from my assemblies and then have the inlet as 2 PSI and the outlet as 1 PSI controlling the pressure drop, using water and monitoring the flow at the 1 PSI point, this in my opinion eliminates the variables from the CV equation and means the flow would be the CV, much like my real tests. However this results in a flow of 210 US Gal/min.

Sorry if I missed any information which would be crucial to solving this but if you have any questions I would be happy to reply with further information, I imagine I am overlooking something relatively simple, but this has lost my faith in the software at the moment.

Cheers,

SteveM

[Boris_M]

Hi Steve,

I'm not an expert about CV calculations as there are I think some methods depending on the fluid and if it is a valve etc. so I'm not sure and there is also the Kv factor. This is not my daily routine so I might misunderstand you.

So you mentioned water somewhere in your post and the CV I know for liquids is Cv=Q*(G/dP)^1/2

Where G is the specific gravity (1 for water), Q is the flow rate (in gal/min I think) and dP the pressure loss (in psi as far as I remember).
So you say, you want to eliminate the square root by making the fraction disappear through causing dP to be 1 as well as G.
I'm not sure if dP is the total pressure drop or the static pressure drop. I would assume it is the total pressure drop.

1.
Did you apply total pressure at both boundary conditions (inlet and outlet)?

2.
Have you checked the dP with goals if it is actually 1 psi or maybe the calculation was not fully converged yet in terms of a flat line? This definition is still an iterative approach and as can, for convergence help, deviate from the defined outlet BC value. The inlet total pressure might be fixed but the outlet has fluctuations as the flow rate is developing and there will be some deviations.

3.
A total pressure-total pressure is not ideal as it leaves the static pressure and the dynamic pressure definition available for changes and the dynamic pressure includes the velocity and therefore the flow rate. Defining the flow rate is a more accurate boundary condition. If you would vary the flow rate to the point where the dP reaches 1 psi, you would see the results should be correct.
You can try this with the parametric study "goal optimization" option and let the flow rate vary between two values to find the goal of the dP to be 1 psi with maybe 0.05 psi accuracy. But you need to consider enough iterations in calculations to find that value accurate enough.

I hope this helps,
Boris