Check Valve Design Reference Material

[bryansonnier]

thread408-257055

I'm looking for good reference material for Check Valve Design. Anybody know of some good reads?

 

[MickMc]

Try this for some basics.

 

[bryansonnier]

Thank you. This was somewhat helpful, however I should have clarified in the subject, I'm looking for Ball Check Valve design material.

 

[btrueblood]

Check this thread: thread408-260735

Past that, what are you stuck on?

Oh, and by the way, no need to post a hello in other forum, this one is on my read list.

 

[bryansonnier]

Cool, thank you.

These check valves are on a PD pump. Basically our seats are bowling out, and our springs are breaking and cupping severly. We are not hearing any complaints from customers and this product has been in the field for many years, except one customer who runs these pumps 24/7. So we started testing again in house and are seeing the issues described earlier.
Springs:
We've tried every range of spring we can think of from stiff to soft, from thick to thin, from large to small contact area, so we are going back to original springs. The cage is such that the ball can't compress the spring to anywhere near solid height.
Seats:
Our seats are wearing out and on occasion sticking once its worn too deep. We are currently using a 17-4 PH H900, although we had tried 440 C but they cracked quickly. Our max pressure can be as high as 30 kpsi.

The kicker is these problems are 99% on the suction side only.

We tried running without springs and saw less wear on the seats but not so much so that it didn't make us worry. The down side though is we did give up a little on the efficiency.

I'll be looking at material selection again but i'm also looking for design information on the seat to maybe choose a better shape and/or seat angle.

 

[btrueblood]

Some questions,

1. "bowling out" should be blowing out? What materials are the poppet/ball and seat made from? Or do you mean the seat is wearing out prematurely? Do you see wear on the ball also? Pitting or other signs of cavitation damage? You ideally want the replaceable element (ball in your case?) to wear first, which means it should be the softer element in the pair...I think. Are the seats replaceable?
2. Triplex plunger pump or other?
3. What is the service (fluid, temperature, inlet and outlet pressures)?
4. What do you mean by the springs are cupping? Do you mean buckling? Or are the spring seats wearing on the check valve poppets? Do the springs have machined seats to bear against, and are the springs ground flat? What material are the springs, and what fraction of running speed is the critical frequency for the spring, and for the spring+ball/poppet mass.

"the kicker is these problems are 99% on the suction side only."

Suction side seats always see a more severe pressure load cycle, when you think about it - from low (suction head) to high (full output pressure) in one half revolution of the crank.

Are you really using a ball as the poppet? What flow rate are you running? Other than really small flow rates, usually one sees a machined poppet with an integral spring seat. How close is the spring to buckling (what is L/Lcrit)?

Sounds like there may be a low enough inlet side pressure that you are seeing cavitation, or incipient cavitation, causing the damage you note (essentially impact damage and/or cavitation bubble collapse, leading to impact damage to the seat and ringing/excitation of dynamic behavior in the spring). You could advise the customer to increase the supply pressure (booster pump), and/or lower temperature, or use a slower pump speed with a larger cylinder/plunger diameter and/or smaller crank displacement to reduce the acceleration head. Harder seats (stellite?) might help.

30kpsi is up there. Our sister company is in process of testing stuff at those pressures, but I have not heard anything from them about pump troubles...though maybe it's their test pump you are worrying about. Realize you are working with a pressure that is a significant fraction of the fatigue design stress for even 17-4 PH, so design of the seat becomes non-trivial. What are the predicted contact stresses on the ball and seat at 30 kpsi load on the ball, and what fraction of yield stress are those contact stresses?

Ok, more than a few questions...

 

[bryansonnier]

1. prematurely wearing, no wear on the ball, no pitting or other cavitation signs, all wear is smooth. Seats are replaceable, mainly trying to keep the maintenance cycle as long as possible.
2. triplex
3. in house testing (where we are seeing the issues personally) water/glycol, 90°F - 120°F, 12-50 psi in, 5000 psi out.
4. springs are wearing out to shape of ball (tried harder spring to keep that from happening and they just snapped), springs are ground flat, they seem to be snapping (if they snap) at the angle transition to flat but they mostly just wear prematurely. i don't remember the critical frequency but we are no where near that.

yes poppet is a ball, possibly looking at mushroom style ball with area for spring to sit but we also like the idea of a moving ball to get even wear.

flow rates for these tests are around 5-7 gpm at 5ksi, and thats another thing we are seeing this wear at 5ksi, usually the pumps don't run at full pressure, they are only rated to that for hydro tests so ramp up and done in a matter of hours or a couple of days. We are doing longevity tests now.

being we are not seeing any complaints from the field (except that one customer that has now gone silent which leads us to believe their issue was in the fluid, i.e. particles) we are not looking to completely change the pump design ,more like tweaking to perfect it.

 

[btrueblood]

Inlet pressure is presumably gauge pressure?

Sounds like the springs are sticking to the ball. I'd machine a spring seat on the ball (making it now a poppet). The wear you are seeing - is the material moving or being eroded away, i.e. do you see mushrooming of the seat material? If it's erosion, how clean is your fluid? Are the seats lapped to match the ball contour, or do you start with a slightly "sharp" edge and line contact between ball and seat? Have you tried matched/lapped fit?

It sounds like the ball may be heavy enough to delay the ball moving to seat, i.e. you end up slamming the ball home, rather than it moving gently to seat at top of pump stroke. Though the same symptoms can be seen when the seat sees cavitation bubble collapse. Lightening the ball (removing material typically, which is why your competition usuually uses a machined "mushroom" poppet, though you could try a carbide ball too) may help.

Is your customer out in the nw corner of the US?

 

[bryansonnier]

The ball may be slamming home, but the volumetric efficiencies of the pump are pretty high, but we may look into a lighter ball or like you said a poppet ball instead of a standard ball.

Thank you for the insight this will give enough to chew on for a while.

 

[btrueblood]

We often confirmed for ourselves that the standard mushroom-shaped poppets will rotate about their natural axis during pump operation (i.e. we marked up the backsides of the poppets, installed them noting the marking positions, then ran the pump, dissssembled and inspected the markings again). Believe it's due to the wind-up and release of the spring, but you should confirm it's happening or not with your valve poppets. If the spring gets jammed against the sides of a ball, the rotation can stop, and wear rates will increase. Rotation is generally a good thing, as it distributes wear more evenly over the poppet and seat.

Today, if I were working on plunger pumps again, I might buy a cheap, waterproof USB camera and mount it in the suction side plenum, viewing the seal end of the poppet/ball. Did this recently for a control valve that was giving us trouble - it worked quite well as long as the water was clear.

We used something like this:

https://www.amazon.com/Crenova-Mega...d=1472136876&sr=8-1&keywords=usb+camera+snake

 

[gerhardl]

I think you should also look beyond the pumps and analyse the whole piping system. Check-ballvalves are the most likely checkvalves to give problems. Common problems (as you know) are uneven load, air in the system, sudden back pressure caused by too high fluid lifting, parallell pumps giving different amounts when pipelines are joined, too high opening pressure against selected spring - then sudden opening and sudden closing again, water hammer caused by anything mentioned above or more.

In my opinion and experience this is more a piping problem combined with a check-valve selecting problem, and more than finding a material solution within the valve.

A soft-closing checkvalve-type might solve your problems.