Clack valves or check valves

[cadser]

Here's one for the smarter chaps among us.

I have read all sorts of threads about the problems we encounter with the seating of ball check valves. I myself have made four new ones for my engine and three are fine and the fourth is a leaker. All made the same way all using the same material - gun metal and steel balls  to seat them and then putting new stainless steel ones in. Base made by D bit. I had never made one and I like the tool it's very simple and gives a nice conical rise to the seat. 

So its a seating problem, I guess if I hit the ball a bit harder I will get a seat, but it will only be by chance and not a skill or repeatable tolleranced method. 

In my professional life I have used check valves that are manufactured by companies like spirax sarco, etc. These are generally flat faced lapped down onto the mating face. They never fail. Steam traps which eject saturated steam condensate (thermodynamic) are made in a similar way and these are so simple in operation and have a loose disk in them that seats on a mating face. So, why do they work so well when they are only two faces that are flat and pressed together? 

I beleive it is two things. First the mating faces are very good finish. The second is to do with the surface area presented to the steam pressure. This is quite large compared with a sphere (ball). I am not clever enough to work out the kind of pressure that will be applied to a sphere covering a hole. As some of the pressure is curving around the ball and not having any effect on the ball, where as on a flat surface the pressure can be calculated and I am sure it will be a lot more than the ball.

your thoughts please?

Brian

[JonL]

I don't know that I could achieve the finish that the professional industry could. Maybe thats part of the difference.

Most diving watches use a screw down case back. It requires good tolerances to get a good seal. In most mass produced industries these days its easily achieved, but wasn't a few years ago. Thats why certain companies were able to charge such a premium.

The russian watch industry came up with a design which suited their production techniques. In this case the back of the watch is not screwed down, but a retaining ring screws down to hold it in place. The caseback sits on a broad rubber gasket. As the water pressure increases the pressure on the caseback increases and tightens the seal further. Still an excellent seal, but more achievable using modest manufacturing techniques.

[cadser]

Thank you for your reply nobby. I though I might have provoked more responses. Or maybe no one can calculate the pressure thats holding the ball down? The ball has only a small amount of area lost underneath it on the low pressure side. Most of the other surface is at boiler working pressure but even not like a flat disk. So my belief is that there's very little effort holding the ball down.

Bri

[uuu]

Take the area of the hole. That's the effective area that the pressure is working on. It's the only part where the pressure above the ball differs from the pressure below it.

But that's the same, whatever the shape of the thing blocking the hole.

Wilf

[cadser]

Hello Wilf. Thanks for your input.

So the area of pressure is very small? If that's the case then very little effort is keeping the thing shut?

Brian

[uuu]

If the hole is small, yes. But the contact area between ball and seat is very small indeed.

Taking the analysis a little further: The forces on the ball should sum to zero, or it would accelerate. Assuming a valve where the ball (or other device) is above its seat - there's a downward pressure on the ball from the steam above it. There are sideways steam forces around the ball, but they cancel out. There's an upward steam pressure on any exposed lower area of the ball outside the seat. Finally there's the force pushing the ball by the seat itself. I'm ignoring gravity here, and assuming no back-pressure.

For a ball of radius Rb, hole of radius Rh, and seat of radius Rs:

Area of top of ball πRb^2
Exposed area below ball πRb^2 - πRs^2
Area of seat πRs^2 - πRh^2

Lets call the pressure of steam P, and the seat pressure, which we want to work out, X

So, for the forces to balance, PπRb^2 = P(πRb^2 - πRs^2) + X(πRs^2 - πRh^2)

Then PπRb^2 - P(πRb^2 - πRs^2) = X(πRs^2 - πRh^2)

PπRs^2 = X(πRs^2 - πRh^2)

X = PπRs^2 / (πRs^2 - πRh^2)

X = PRs^2 / (Rs^2 - Rh^2)

Feeding in P = 100psi, Rs = 0.09625" (5 thou over a 3/16 dia), Rh = 0.09375 (3/16 dia)


Then X = 1850 psi if my thinking is OK. Force between ball and seat = PRs^2 = 2.8 lbs

I've not double-checked this, the weather is too inviting outside.


Wilf

[pault]

A problem I have had is where commercial balls are actually lobed (where the diameter measures the same in all orientations, but the centres differ, if that makes sense). This will result in a valve that will seal perfectly sometimes and leak like a sieve other times. It just depends how the ball orientates itself.
There are two basic styles of valve, metal to metal and valves which use one or more compliant components.
“Plate” style valves are a little more work to make when metal to metal but do work very well. However for whatever reason model engineers tend to avoid them.
With “ball valves” no matter how good your workmanship is you are reliant on the accuracy of the balls, and as I have said they are not always as good as they should be.
Valves with a compliant element when properly designed tend to work very well as long as they are made from materials which are happy in the environment they are being used in. A typical problem is the use of nitrile balls in clacks. Nitrile is rated for use at about 120°C continuously and about 150°C intermittently. Steam at 90 psi is at 166°C so nitrile will tend to have a relatively short life. In my experience it tends to go hard, deform and crack after a time. Not a major problem as long as you replace it before it breaks up or disappears down the pipe.

[Jim]

Where the design permits I make my ball valve seats as a separate screw in item to allow me to get a good true surface for the valve seat free of any marks that would effect the ball sealing. As already mentioned there are other factors in play that need to be taken into account but considering the numbers of locos we see steaming and running on club tracks then the vast majority of ball valves work well provided the basic parameters are followed regarding ball to seat diameter and finish.

Jim.

[cadser]

Thank you for the replies. I have in the end changed all the balls for nitrile and when the Viton ones arrive I will pop them in. On and off I have been making quite a lot of check valves. I think my bronze supplier is quite happy. Some are a little larger and these are on the back head. So this morning in readiness for next weekend bank holiday I steamed my engine. I still have lots of water coming out of the drain on the injectors. Something I have lived with! I just assumed I always had enough steam with half a turn of the globe valves. With the valve fully open the drain stopped leaking 👍. If I backed the water off I could get the same result. All in all I am more confident that I am getting there with the injectors,water feed and the check valves holding.

I also fitted filters to the feeds which is something that the tender did not have. The builder used fibreglass resin to seal the tender after building it. It does not look pretty but at least it does not leak. Maybe in time I will make another tender? But for now I am happy.

Wilf I left school with no certificates so your equation is lost on me, but thank you for doing it.

Brian

[cadser]

What size are your water feed pipes and outlet pipes from the injectors to the boiler?

Mine are all 3/16 and i think they are too small

When i turn on the water there is a dribble at the injector overflow then turn on the steam. I can hear the water squeezing into the boiler like a fizzing. However i still have flow at the injector. If i throttle back the water the overflow turns to steam and stops working. Its so frustrating.



Bri


Read more: modeleng.proboards.com/thread/9425/clack-check-valve-article-drawing?page=2#ixzz5myt6bzes

[gwr14xx]

I think you may have a minute air leak on the water feed. When I first got my 14XX, I modified the pipework with a 2 way valve so that I could select water from the side tank or the tank on the driving trolley. No matter what i did, the injector always spilt more than it injected. Having been advised that it may be an air leak, I swapped the 2 way valve for a plain 'T' piece, and made a crude cap (short piece of rubber tube with a ball bearing shoved up it) to go on the leg to the driving trolley when using just the side tank - Hey presto!, the injector worked perfectly (and still does).

Regards,
Eddie.

[cadser]

Hello jim. There may be some mileage in the air leak. I will investigate this week when i get home.

Thanks for the pointer.

Brian

[cadser]

Ok. So when I got home I stripped the running boards off the engine and the cab. Sounds harsh but I do not have them fitted fully as I still have detail work to do on them. This allowed me to get at the water valves which are under the cab floor. I can not see any leaks and all the joints look good. The glands on the stems were ok so if air has been the problem I don't know where it's getting in.

I had a quick look at the web site that covers the build of a modelworks engine. The tender he made had the water valves in the tender next to the hand pump. These will be submerged, so if the gland leaks it does not matter. It also cuts down the fitting between the injectors and the feed source and has them handy. Mine are very low in the cab and I have to reach over everything to get at them. So it would be nice to have them easy to get at. It also means that they can be oversized with larger pipe running to the fron t of the tender undearneath out of sight. This will help with flow rates. It will take me a day or two to do the mods as I will have to make the valves.


I forgot to mention.  I ran for 2.5 hours with the nitrile balls in the check valves. I had a look at them and they were just as I had put them in and they worked great.  Polly models sell them and they state they will last for a long time so fingers crossed I can leave them alone and change at the end of the season 


Bri

[Jim]

That sounds like good news to me Bri. Well done, now to get mine sorted.

Jim

[cadser]

For those interested and if your like me need help with things i have just come across John Baguleys excellent web site. He is on with a re build of a Winson Britannia and is correcting a lot of issues.

I was led to it via Wikipedia whilst looking at LBSC’s page.

Bri

[cadser]

It has been a while since i posted, work is very busy and we have been decorating. I have had a little time in the workshop, putting a more scale ducting on the left hand side of the boiler, the Winson one was way over scale. I also made a start on the twin steam valve on the left hand side just in front of the cab. That should be finished soon.

I also decided to buy the drawings from steam workshsop. They are every well done and give me loads of information i did not have. I have drawn up the reverser to replace the rectangular block that Winson provided with the the model.
I managed to get a work wheel set and a set of bevel gears that will do nicely. I had to wait a while ass they came from china.

That's all for now, stay safe and away from this awful virus.

Brian

[cadser]

Today was glorious weather so i decided to have a steam up. A good chance to test my new rolling road. I have always thought that the timing of the return crank was a bit out? When starting the engine i have to open the regulator fully to get things moving, so being able to start and stop to move it a little and then run again is a bonus.

The left hand one needed about 5 degrees and the right hand one a little more. It defiantly improved the running.

If i put the engine in mid gear and open the regulator i get steam from the chimney, is this correct? Looking at Mr Baugleys improvements to the Winson he fits o rings to the valve liners, this looks a good idea, but only if the things are leaking? It a lot of work if they are not leaking? I will ponder over this change whilst i make the steam drains for the cylinders as the mechanical ones are rubbish.

Bri