True formula for flow coefficient

[SpudUke5]

Well i dont know how to use GGDT, sad i know

I have been informed that there is a page on GGDT website that can help but i havent taken a look at that.

But thats not what i am here to talk about.

Since i want to use GGDT, i need to find out how to get a flow coefficeint, and i have done some research and have seen this formula: Cv = K * D^2 which was posted by ragnarok here.

But as he mentions in the topic, that is not the correct formula for Cv. So, i went searching on the internet to find the correct formula. As a result, i found this formula (scroll down to the Gas flow formula, first formula).

So i am wondering how would i figure out this formula, because i am kindof bored, and i am learning Java this semester, so i want to make a program that gives you a flow coefficient.

So can one of the knowledgable engineers here help me out with better understanding of the variables in the formula? (which the variables are defined next to the formula)

EDIT: there are also tables that show Cv's, but im not sure if those can be applied to the valves we use.

[clide]

Well, the only actual way to get Cv is to measure it.

However, now that GGDT has the efficiency setting for flow coefficient you don't really need to find the Cv. You just need to know your seat diameter and the efficiency for the kind of valve you are using.

Here are some estimated values for different valves. The exact value will change depending on the specific valve, but based on what I have measured these should be in the ballpark:

Sprinkler valve - 30%
Coaxial piston valve - 35%
Tee piston valve - 40%
Burst disk - 60% (?)*

*I'm not really sure about the burst disk one since I haven't ever done measurements on one.

[MrCrowley]

Got anything for QEV's clide?

[clide]

From what I've seen QEV's can be very different in their flow paths so I'm not sure if I can generalize them. I might try to dig up some data, but your best bet would be to look for spec sheets since flow coefficients are often measure by the manufacturer and would be far more accurate than a generalized number.

My WAG before looking up anything would be 35%

Edit: Well after looking up the manufacturer specs on a QEV that I have it either confirms that they can't be generalized or disproves that the manufacturers data is accurate. Based on the seat diameter and the Cv it would have an efficiency of 22%, however I think that is most likely because the fitting where you hook up the chamber has a smaller diameter than the seat of the valve meaning flow is probably choked there rather than in the valve.

My research did confirm that it is quite easy to find the Cv for QEVs so I would still go with my earlier suggestion of just finding the specs. If you absolutely can't find them then I would say 35% wouldn't be a bad estimate.

[boilingleadbath]

The formulas on that website are for figuring out <i>flow</i> given <i>Cv</i>. Since the GGDT already does those calculations, you don't need them.

They have a table of CV values at the end of their .pdf, but that won't do you any good... you don't even know their valve geometry, and I'm sure you don't want to pay for their fancy gate/ball valves!

[POLAND_SPUD]

hey why don't we try to put here (or in a new thread) flow coeficients of the most popular valves ?

I can't find flow coef for the QEV that I have but I noticed that this model is the one most commonly used here

[akb]

I designed one launcher using a Deltrol 3/4" QEV and their published Cv feeding a tennis ball barrel (SDR21 2.5). Actual performance was about half of predicted, as I recall (it was a very long time ago). I suspect they measure Cv for exhausting an air cylinder (their intended application), and this is not quite comparable to feeding a barrel with the flow. Or it's just marketing, multiply actual measurements by some factor.

I've had the best luck building a launcher, measuring the velocity, and using that to calibrate the flow efficiency. Then changes to the launcher can be predicted within some range. Unfortunately changing chamber volume seems to change valve efficiency, so you can only predict performance over a limited range.

And yes I know this is an old thread I'm responding to.

AlanB