Safety valve

[mcr]

I have made a safety valve following the typical design but I am now having problems getting the ball bearing to sit correctly.
ANY ADVICE?

 

[RICHARDDV]

HERE IS A TRICK I USE WHEN WORKING ON HYDRAULIC CHECK VALVES. GET A NEW STEEL CHECK BALL THE SAME SIZE AND INSERT IT INTO THE BORE SEAT AND USING A BRASS DRIFT LIGHTLY TAP THE BALL ON THE SEAT . REMOVE THE STEEL BALL AND DISPOSE OF IT. THEN INSTALL ANOTHER NEW STEEL OR BRONZE BALL, ASSEMBLE AND CHECK FOR LEAKS. THIS METHOD WILL TAKE MOST IF NOT ALL THE MACHINING IMPERFECTIONS OUT OF THE SEAT . GOOD LUCK
RICHARD

 

[GWRdriver]

MCR,
Depends upon what you mean by "seat correctly." If by that you mean that it will not lift and then re-seat within a reasonable range of pressure without losing too much pressure, and the spring tension is correct, then the problem will be in the seat and ball chamber geometry. If you mean that when seated there are leaks around the ball then the first step is to insure the hole is as round as possible which for us means reaming or boring rather than drilling. Opinions differ on burnishing a safety valve seat as you would a check valve. Whose or what design did you build to?

 

[mcr]

I have had another go at the safety valve much better results still slight hiss as the boiler heats up loss of steam is small.
The valve blows at 42lbs consistently which I am pleased with shuts back at 40lbs but the valve still has a slight loss of steam.
I reamed the hole and seated the ball on the hole giving a slight thump.
Throwing away the ball and replacing with a new one.
Any further advice would be welcome.
Mark

 

[GWRdriver]

Mark,
One of the recurring problems reported by builders of model safeties (aside from weeping seats) is delayed closing, where too much pressure is released before the valve closes, and this is directly related to chamber geometry.

For larger model boilers and higher pressures, say 80-90psi or above, a 2lb differential could be a problem because when the boiler is being worked the safety might need to open constantly to drop 2lbs when fewer releases with larger pressure drops per release would be better, and safer. For smaller engines and stationary plants at lower WP (working pressures) the 2lb (or thereabouts) differential is desireable so that too much available pressure isn't released at one time. Either way your differential it indicates that have your chamber geometry just about right for a small boiler.

I have a small scale locomotive which has a single ball-seat safety which is so pressure sensitive and well balanced its releases as it goes along sound like Pip . . . pip . . . pip . . and I love to hear that little guy. My guess is the pressure differential is a matter of a few grams and is too low but I still like to hear it anyway. Also, because prolonged releases rarely occur, this indicates there is very good balance between firing rate and steam consumption.

 

[xo18thfa]

.
.
.
I have a small scale locomotive which has a single ball-seat safety which is so pressure sensitive and well balanced its releases as it goes along sound like Pip . . . pip . . . pip . . and I love to hear that little guy.
.
.

Harry: Post a drawing of that valve. Pppllleeeease

 

[GWRdriver]

Bob,
This particular one is an Aster Models product so no drawing is available but it's behavior is typical of many Aster safeties although individual behavior varies from valve to valve. They have pretty much got tiny safety valve proportions nailed down. I've been tempted to pull a wax core of the chamber to see if I could replicate the geometry but I haven't bothered to do that yet.

I sound like a broken record but the secret is the ball seat and chamber proportions which need to be accurate and consistent in order to produce the crisp and predictable behavior we look for. Spring strength and tension are obviously important also and that presents us with a lot of variables which if just a whisker off won't produce good results. I tried building a series of small ball valve safeties a few years ago with generally disappointing results and a major part of the problem was I was working with out of date seat and chamber design information. There is now far better information available. My current resource is "Safety Valves for Model Locomotives" By Gordon Smith, Engineering in Miniature, 2001, and no doubt there have been improvements on this information.

 

[gbritnell]

There are basically two types of safety valves, the plain spring, ball type and then the spring valve type.
The spring, ball type are nothing more than a spring loaded ball sitting on a seat. When the pressure gets high enough it will lift the ball from it's seat and release steam. It doesn't have a good range of action because of it's design. The most important thing is the seat. It should be cut with a cutter (home made) that will give a reverse edge to the seat as opposed to a flat surface. Once machined it will only require a slight tap on the ball to seat it into the sharp edge. This will make it seal the best.
The spring, valve type is made much like their larger brothers. It has a valve face that sits upon the seat face almost like an internal combustion engine valve (45* seat). Above that seat is an enlarged flange which is still smaller in diameter than the full bore of the release area. How it works is this: when the steam pressure rises to the point that it lifts the valve from it's seat the pressure is now applied to the enlarge flange which then increased the pressure against the load spring. This causes the valve to 'pop' open. When the pressure has decreased to the point that the valve started to close the reverse happens and the valve drops tightly against the 45* seat. These types of valves can be adjusted from the top of the valve by turning a slotted sleeve which in turn puts more pressure on the spring. They work quite nicely but are picky to build because it takes some adjustment of the flange diameter to get them to reset properly. Not that they won't reset but rather in how much steam pressure is dropped before they reset.
gbritnell