There is no simple answer to the question about closing times because there are so many factors which influence the movement: localized pressures, viscosity of fluids, temperatures, constantly changing spring rates, etc.  We simplify the problem to a basic spring-mass acceleration problem.  This is highly oversimplified, but it will at least give you a "worst-case" (fastest time) scenario.  In general, spring assist will help minimize water hammer.
I do not know of any sites that compare the behavior of the various kinds of valves.

I am not aware of any sites that publish that kind of information.

Here are some references that might be of help, but you would need access to a university technical library to obtain them.

GA Provoost. “A Critical Analysis to Determine Dynamic Characteristics of Non-Return Valves.” 4th International Conference on Pressure Surges, September, 1983.

ACH Kruisbrink. “Check Valve Closure Behaviour; Experimental Investigation in Water Hammer Computer Programs”. 2nd International Conference on Developments in Valves and Actuators for Fluid Control. Manchester, England, 1988.

ARD Thorley. "Check Valve Behavior Under Transient Flow Conditions: A State-of-the-Art Review." Vol. 111, ASME, June, 1989

Note the dates. The interest seems to have waned. Power plants (nuclear in particular) were addressing these problems quite a bit. Waterhammer issues were significant. They are still around obviously, but the understanding and changes to operating procedures and/or designs helped to avoid the problem.

While you can’t guess the closure speed, you can make some qualitative assumptions from the design. If (1) the disc is light weight, (2) closure is assisted by springs, and (3)the full stroke is short, then you will have fast closure, which of course is what you want. Swing checks do not meet any of those criteria while nozzle checks meet all three. Tilting disc, duo/double door, and lift (piston-type) check valves fall in between.

The only accurate way to predict the accurate closing time and corresponding pressure surge magnitude of a specific check valve design is to use dynamic performance curves.  These curves are specific to check valve models but can be used for different size valves with the same approximate geometric ratios.  These test are very expensive to conduct therefore most manufactures have not done much testing.  I have curves for a specific type of swing check and all large bore Nozzle Checks.  I have used these curves to accurately predict pressure surge +/- 20 psi.

Try

"Dynamic Testing of Check Valves" William Rahmeyer
www.engineering.esu.edu/Departments/cee/faculty/rahm/check.html

Also, there was some fundamental research doone in this area by Delft Laboratories (Netherlands) which I am also trying to find a copy of.

The best way (only way really) is to use test data.  Otherwise you can predict all sorts of really high pressure surges and piping loads which may or may not (more likely) exist.

ASME publication 82-PET-16, titled, "A Facility and Approach to Performance Test of Check Valves" specifically address water hammer as they relate to check valves.  This publication is probably not available in print.  I would be happy to fax it to you.

jdptx

This sounds like an interesting publication.  If possible, could you fax me a copy of it?  

Thanks

Bob S.
315 349-1836 (Fax)

I had some experiments in swing check valve included systems. Fast closing gate valve assumption gives approprite results. Some equations given in the literature are difficult to use because of unsufficient knowledge about the check valve structure.    

jdptx

I would love a copy of ASME 82-PET-16.  Please fax a copy to Brian L. at 570-752-4962.

Thank you

I have an article that was persented ath the 5th International Conference on Pressure Surges in Hannover(1986). It covers the dynamic characteritics of check valves and a possibility to "scale" the dynamic characteristic into a dimensionless form allowing a range of geometrically similar check valves (with differnt spring loadings) to be presented in one dynamic characteritic curve.
The article was written by Mokveld Valves, the Netherlands and Delft Hydraulics Lab.

E-mail me for a copy

Hi Mokv

Could you please end me a copy on blenrayaust@yahoo.co.uk

Cheers

Stanier

Sharing knowledge is a way to immortality

We have approximated check valve closure time by assuming the check valve disk moves at the fluid velocity.  To apply this method, simply divide the distance the check valve disk must travel by the reversed fluid velocity.  If you don't know how far the disk must travel, you can assume the distance is on the order of one pipe diameter.

The difficult part of this approach is determining the reverse fluid velocity as it will be changing over time.  Once you calculate the fluid velocity as a function of time and approximate the valve closing time, then you will have a value for the fluid velocity at the time of valve closure that you can use in a waterhammer calcualtion.

P.S. Mokv - I am interested in the article you mentioned.  Please e-mail me at elicson@fauske.com.  Thanks.

Is the pressure surge a field artifact or caused by pump station?  There is a huge difference in how you deal with the surge.

By my experience most surges are field artifacts.  If so then very little can be done at the pump station about that.

If pump station caused it is usually a shut down problem.
Silent check valves are called silent because they perhaps can stop the hammer caused by pump shut down.  Surge anticipators may help also with shut down.

PUMPDESIGNER

MOKV:

I would appreciate if you could also e-mail me a copy:

mtna@ramboll.dk

Best Regards

Morten

Hi JDPTX & ROBV!
I'll like to get the copies of what u got. Plz email me at n_nasir@yahoo.com.
Thanks & regards

Another interesting reference that you might consider faxing to all of those requesting copies is

"Hydraulic Hammer Theory and Appplication"
PVP-Vol.278 ASME, pp.67-83 (1994).

NFS Safety Notices No.98-02(November 1998) dealing with Water Hammer.

The issue is not so much the exact closing time of the valve but its relation to the critical time for the pressure wave.