6 replies to this topic

[Brightonnk]

Hey Everyone,

Im doing a project on ball valve cv values and the range of products I have use a spherical shell ball (also known as hollow ball) for the larger sizes. From the mnufacturers cv values I've noticed that when using hollow balls the cv is considerably more. So I did a few quick calculations on a DN80 hollow ball valve (assuming it is full bore) and the cv value was a 1/3 of the cv value stated in the brochure...

I was under the impression that the same valve would only vary slightly if it was a hollow or solid ball.

is it normal that my cv values are so different? or have a made a mistake in my full bore calculations. I am aware that slight turbulence will occur in the hollow ball but thought it was negligible.

Any advice would be appreciated

Thanks

[kkala]

Not having experience on hollow ball valves, I had this impression too; but then I have looked at following
http://www.boehmer.d...ish/widerst.htm
http://www.associate...apollo-tech.pdf (page 3)
http://www.eng-tips.....cfm?qid=267573
So it seems that a hollow ball valve presents higher resistance to flow (hence lower flow coefficient Cv) in comparison to a solid ball valve, both full bore.
Nevertheless it is noted that frictional ΔP of such (full bore) valves is quite small when fully opened (hollow or solid ball). And ball valves are on/off valves. So the difference in ΔP / Cv is not significant in practice.
A more clear explanation on mentioned "higher resistance to flow" (than the one in the third link) would be welcomed at any case.

Note: ΔP of reduced bore ball valves is mainly due to flow contraction and enlargement.

Edited by kkala, 04 January 2013 - 01:56 PM.

[Art Montemayor]

brightonnk:

My advice is still the same as in: http://www.cheresour...ow +ball +valve

Pick up the phone and obtain the Cv from the manufacturer.

[katmar]

It is not unusual for low quality valve manufacturers to overstate the Cv's for their valves. There is a strong correlation between the reliability of the valve and the reliability of the information provided. For the reliable valves like Fisher, Masoneilan and other well known brands you can stake your engineering reputation on their valve data, but if the information seems dodgy I would be suspicious of the valve too.

[kkala]

Let us consider a ball valve, 100% open, on a 3" Sch 40 pipe to transfer water at 1.5 m/s (flow rate = 25.8 m3/h). Frictional losses through a solid ball valve, full bore, can be K= 0.07 velocity heads (as understood from 1st link of post No 2). Looking into Table 6-4 of "Additional Frictional Loss for Turbulent flow for fittings and valves" in Perry (7th edition, similar table in other editions), K=0.07 seems reasonable for the case.
Considering several ball valve cases in an attempt to clarify situation, frictional ΔP through a ball valve can be as follows.
1. Full bore, solid ball valve : ΔP=0.07*1.5²/(2*10) m H2O = 0.0079 m H2O (g=10 m/s²).
2. Reduced bore (80/65 mm) solid ball valve : ΔP=0.0574 m H2O (K=0.51, 1st link, post No 2); that is 7.3 times more, so Cv would be ~2.7 times less compared to 1st case. Check "assuming it is full bore", post No 1.
3. Full bore, solid ball valve, slightly misaligned due to imprecise construction. ΔP higher than 0.0079 m H2O.
4. Full bore, unusual "hollow" ball valve, ball empty sphere with two opposite orifices on the shell; like the one referred in http://www.cheresour...cv-calculations '> http://www.cheresour...cv-calculations . ΔP higher than 0.0079 m H2O, due to some flow enlargement - contraction etc.
See http://www.eng-tips.....cfm?qid=267573 '> http://www.eng-tips.....cfm?qid=267573 , especially post dated 19 may 2010.
5. Full bore, hollow ball valve, ball shell with tube (equal to pipe dia) welded across it. Probably brightonnk speaks of this valve, which needs confirmation. ΔP ≥0.0079 m H2O. Flow "sees" same path as solid ball valve, yet construction precision can be less. Or valve may be slightly "missfigured" during operation, resulting in somehow higher ΔP apart from leaks. For best quality valves of this type ΔP=0.0079 m H2O.
Conclusions based on above, full bore valves 100% open.
ΔP of above cases is quite small. Precise experimental ΔP is doubtful, and so is resulting K or Cv. Differences in corresponding Ks are even more doubtful, so conclusions can be uncertain (even if measurements by supplier are "honest").
Note: K of ball valves do not have practical significance; similarly to http://www.cheresour...-through-a-tee/ '> http://www.cheresour...-through-a-tee/ . So there may not be interest for precise measurements of K (and resulting Cv).
Nevertheless Cv of hollow ball valves is ≤ Cv of solid ball valves, according to interpretation as above.

[Brightonnk]

kkala: In your first and last post you said the cv value for hollow balls are less due to higher resistance this makes sense to me. In your first post 3rd link this clarifys my thoughts on using hollow ball in the first place. I though the only benefit was cost adn weight and as they're only usually used in the bigger size valves- using hollow balls seem appropriate. Thank You for the break down in your last post- you've explained it in a logical way and so the conclusion is that i can assume difference of cv values are small.

Art:As I have explained in previous posts, I am working WITH the manufacturer as part of my uni course. they have emplyed me to calculate and then later test their entire valve portfolio- so in hindsight i will get exact values of the flow coefficient, however as part of my project i have been asked to calculate/estimate the cv values to help me in becoming an engineer. As i am only at uni the company thougth this type of theoretical work would help me apply maths/physics to real life problems.

So yes I could ask the manufacturers but they are unsure themselves about their values hence why I am doing this project.

Katmar: I think you're absolutely right - the reason that I am doing this project is because the technical data the company have is unreliable as the source of the information is unknown.

Thank You for your replies

[kkala]

In addition to the above, I think experimental determination of ΔP ~ 8 mm H2O on a valve located at centrifugal pump discharge is "next to impossible" for even a good lab, since pressure variations will be much higher. On the other hand measuring total ΔP of numerous identical valves in series will create issues, e.g. which ΔP of interconnecting pipes, not well developed flow regimes upstream / downstream each valve. Or flow velocity is increased above recommended values to result in measurable ΔP across the single valve? Any advice would be welcomed.
It is noted K=0.05 is determined in attached "FBball.xls", indicating variation of K as solid ball valve is closing.

Attached Files

•  FBball.xls   24KB   17 downloads