Ball-type Check Valve resistance coefficient (K)

[gibsi1]

Having an impossible time tracking down a resistance coefficient for a 2" ball-type check valve. Can anyone help me out?

 

[vanstoja]

Blevins,R.D., 1983, "Applied Fluid Dynamics Handbook",Van Nostrand Rheinhold gives in Table 6-10, case 9 for a ball check valve
K=2.3 (From Crane Technical Paper #410)
K=0.5 (From Boston Soc. Civil Engrs. V42#3,pp.263- 286,1955)
Idelchik,I.E. etal,1986,"Handbook of Hydraulic Resistance",2nd Ed., pg. 429 gives the following formula for a spherical valve with a spherical seat within 0.125<h/D<0.4
Lambda=K=2.7-(0.8/(h/D))+(0.14/(h/D)^2)
D is the valve diameter and h is the valve opening distance not defined for a spherical seat in a spherical valve.

 

[gibsi1]

The Crane book does not have ball check valves in it. Neither does Cameron Hydraulic. As far as the formula you listed, I'll see if I can track down the valve opening distance. This system is already in service, so I don't have the valve or it's information in front of me.

 

[itdepends]

Extract from Fluid Flow Design Software version 1.7.

The L/D for a ball check valve is 150



For each pipe diameter, D4, D3, D2, D1 (exit to entry)
Calculate the fluid velocity,
V = Q / A
Calculate the Reynolds number,
If mixture then, Re = ?mixVD / ?mix
If vehicle then, Re = ?vehVD / ?mix
If liquid then, Re = ?liqVD / ?mix
Calculate the Fanning friction factor, ƒ
If Calc method = Newtonian mixture, vehicle, liquid then,
If Re < 2100 then, ƒ = 16 / Re
If Re > 3000 then, ƒ = 0.0625 / (LOG(k / 3.7D + 5.74 / Re0.9))2
If Re > 2100 and Re < 3000 then, interpolate
If Calc method = other procedure then,
Calculate ƒ using the selected procedure (eg Lazarus & Neilson)
Calculate the head loss in metres of mixture,
?Hpipe = (4ƒL/D)V2 / 2g and ?Hfittings = KV2 / 2g
Calculate the pipe wall pressure drop,
If mixture then, ?Ppipe = ?mix g ?Hpipe
If vehicle then, ?Ppipe = ?veh g ?Hpipe
If liquid then, ?Ppipe = ?liq g ?Hpipe
Calculate the fittings pressure drop, ?Pfittings = ?mix g ?Hfittings

Have fun

 

[LHA]

I've used 2.5 for swing check valves in the past, don't remember a source. I would imagine it would be similar, maybe a little less.

K = Darcy's f * Leq(ft) / D(ft)

Leq. = 19.0 for a 2" swing check.

I've probably been no help, but...

 

[gibsi1]

I found a correlation between K and Cv. The manufacturer listed Cv on their website, but not K.

K =~ 891*(d^4)/(Cv^2) ... with "d" in inches

Using the Edward-supplied Cv value of 51.5, and an ID of 2.067 (2" SCH 40), I came up with 6.1323. Given that similar to various lift check K values, it'll do.

Thanks for all the advice! Itdepends, your reply was very informative, though a little more involved that I need. Thankfully, I'm not designing a system. I'm just trying to determine if we need some piping or pump modifications to better balance one of our sump pump systems.

 

[recoveringEngineer]

Lots of good stuff in the above posts. My older (1969) Crane TP 410 lists an in-line ball check valve as L/D = 150 and convention check valve as L/D =135 provided both are FULLY OPEN.

Iha, very insightfull. For complete turbulance (rough pipes) f = 0.019 for 2 inch sch 40, therefore, K = 2.9 for the ball check and 2.5 for conventional swing check, your number.


My newer Crane TP 410 (1978) lists K = 100 * ft for a swing check with ft = turbulent friction factor for 2 inch pipe = 0.019 so K = 1.9. For 50 gpm, L/D = K/f = 1.9/0.023 = 83 diameters. It omits the in-line ball check valve.

gibsi1, IMO, k = 6.1 seems a little high for a ball check.

It is interesting to note Crane (1969) suggests a minimum pressure drop across the ball check valve of 0.25 psid (horizontal) and 2.5 psid (vertical) to fully lift the ball.


Donald Blachly, PE

 

[gibsi1]

Very good info. If a ball check is truely K=2.9, why do 2" lift-type checks range from 1.1 to 11.4? I was under the impression that ball checks were highly restrictive.

 

[recoveringEngineer]

With that much variation, there must be pretty significant geometry differences. You should check with the manufacturer.

I found Cv of 107 and 130 so there is a lot of variation.

But your Cv is from your manufacturer.

Where did you find the relationship of Cv to K?

Donald Blachly, PE

 

[gibsi1]

In the user's manual for the PIPE-FLO software. Is Cv dimensionless?

 

[Latexman]

Perry's 6th Ed. says K = 70.0 for a ball chyack valve.

Good luck,
Latexman

 

[gibsi1]

70?! Well, I did stumble across a random website that said K=65, but I didn't know whether to trust it.

 

[gibsi1]

Found another reference for ball check valves...

http://www.haestad.com/library/books/awdm/online/wwhelp/wwhimpl/java/html/wwhelp.htm

This books lists a ball check at a Kl of 4.5

 

[LHA]

At k=70, the minor loss at EACH check valve would be:
4.3 feet (or almost 2 psi) at 2 fps;
27.2 (12 psi) at 5 fps and
108.7 ft (47 psi) at 10 fps.

You would need a booster pump at each valve!

Further, the k of a 3/4 closed gate valve is 24, so a ball check would be almost 3 times as restrictive as a 3/4 closed gate valve. I've always used gate valves for isolation; I should start using ball check valves ;-0 Just a little weekend humour.

I am fairly certain k is nowhere near 70 for a ball check.



The Chinese ideogram for “crisis” is comprised of the characters for “danger” and “opportunity.”

 

[gibsi1]

Any opinions on using the "calculated" 6.1 vs. 4.5 from the book listed above?