How An Injector Works

[paralleler]

While going through old CallBoy newsletters of the Golden Gate Live Steamers, I came across this article. Originally published with permission of Dick Bagley of the Riverside Live Steamers, this article was found in the March & April 1976 issues. More excuse to thank our parents, grandparents, ect. for the efforts put into developing things like this. I added a good diagram of an injector I found on the Internet, possibly Anthony Duarte of Eccentric Engineering would consider updating the diagram with the cross section of one of his injectors.

Happy New Year, Bruce & Harley

[AnthonyDuarte]

This is a very nice explanation on what happens inside an injector that still focuses on the most important element: momentum. Too many explanations get caught up in principles like Venturi, Brenoulli, thermal expansion, etc... all these things are happening but do not explain why it is able to overcome boiler pressure.

I would like readers to note that Bagley's excellent explanation ends at the delivery nozzle. A lot of folks are under the impression that velocity is critical throughout the entire length of delivery pipe, and this is why it can't have any sharp bends, but that's not really the case. If you don't believe me, consider how many countless full sized injector designs have an immediate 180 turn followed by an immediate 90 degree turn in the injector body itself (including Ohlenkamp, SuperScale, and EE).

As far as velocity is concerned, its work is pretty much finished once the water leaves the delivery nozzle. The purpose of the delivery cone is to convert that momentum and velocity back into pressure, not to make the steam/water mixture go fast. Delivery lines are very forgiving as long as the inner diameter of the pipe and fittings is sufficient.

If you've got 150psi in the boiler and put a pressure gauge on the delivery line, you'll see the delivery pressure grow to a few PSI over the boiler pressure and settle back down to 1-2psi above once the check valve opens and remain there while the injector is operating. If you slowly close off the check valve, you will see the pressure continue to clime to 200psi+ before the injector breaks.

One last comment on one of Bagley's statements regarding the gap between the steam cone and combining cone:
"The mouth of the funnel must not be too large or too much water will enter and swamp the jet. If too small there will not be enough water to condense the steam."
Too much water entering the jet is not a bad thing. If you've used an Ohlenkamp, SuperScale, or EE injector, you've noticed this phenomenon regularly. When you start the injector, water gushes out the overflow until the water valve is cut back because too much water is entering that gap. Restricting the water valve is what regulates the flow to that sweet spot between too much water and not enough. Putting that ability in the hands of the fireman is (in part) what gives these injectors such an enormous operating range (35-160+). If the gap was sized for exactly the right amount of water at a given pressure, you end up with the problem that plagues so many injectors out of the UK. They hit ~100psi and just stop working.

Thank you for sharing this!
Anthony

[Berkman]

When you throttle back the water supply, does it negatively impact operations or just heat the water more entering the boiler since using the same steam with less water volume?

Also remember someone on here used smokebox preheater loops in conjunction with an injector, which I had always thought was a no no, but pretty interesting to squeeze more heat out of the injector system, and improve the already pretty good efficiency.

[AnthonyDuarte]

When you throttle back the water supply, does it negatively impact operations or just heat the water more entering the boiler since using the same steam with less water volume?

In Bagley's paper he mentions that the momentum out of the delivery cone is notably higher than the momentum required to overcome boiler pressure. When you cut down the water valve you are also reducing the momentum out of the delivery cone, since the lbs/s of water included in the total momentum is less. You will only be able to cut it down until the momentum out of the delivery nozzle is no longer enough to overcome boiler pressure, or until it is not enough water to sufficiently condense the steam back into water.

All that to say there is no "negative effect" of cutting back the water, at least not an effect that I would refer to as negative. There's just a range in which the injector will still work. One disadvantage of having an overly restrictive delivery line is that you will have a much narrower range in which the injector will work, because more momentum will be required to overcome the restrictions.

Your observation with heating is correct. When you reduce the water, less water is being delivered while the same amount of steam is passing through the steam cone. So you're putting in less and hotter of water in your boiler. Depending on what you're doing on the track, this can actually be a major advantage. For some actual numbers... In my tests, at full capacity (water valve as wide open as it can be while still allowing the injector to operate), the delivery temperature is around 130-140F. When the water is cut back, the delivery rate is cut back as much as 65% of full capacity, and the delivery temperatures are closer to 160F.

The D&S has their injectors (technically inspirators) running almost constantly on the climb to Silverton, but cut back as far as possible so they're drip feeding the boilers with hotter water than if they were running them full bore in cold bursts.

Also remember someone on here used smokebox preheater loops in conjunction with an injector, which I had always thought was a no no, but pretty interesting to squeeze more heat out of the injector system, and improve the already pretty good efficiency.

The only problem I can think of is if you have superheated water dripping back to the injector. If you have a check valve before and after the heater coil there won't be any negative impact on the injector.

[Fender]

Anthony, a friend has a Superscale standard injector mounted high on the side of his boiler, and it works very reliably in this position. I want to move my injectors similarly, because this prevents water from siphoning out the overflow, if a small hole is drilled in the overflow line right below the injector. Also, I wouldn’t have to close the water valve on the tender when shutting off steam to the injector. My only concern is that I may still need to fiddle with the water valve due to varying boiler pressures?

[Berkman]

This seems like a lot of work for little to no gain, you will still need to throttle the water with varying boiler pressures. Keeping them mounted low means that getting water to the injector will never be a problem and in the event they were to get too hot etc, having them mounted in a non lifting position means you can flood them with water and cool them off.

Basically the majority of later steam power favored non lifting injectors. Even 4501/630 had their lifting injectors replaced with non lifting during their restorations. I'd only go with a lifting arrangement if it was a necessity due to placement or your prototype had the injectors in front of the cab as a characteristic look with control rods running back into the cab. But you'd want a nathan simplex style injector etc to make this look right.

[AnthonyDuarte]

Anthony, a friend has a Superscale standard injector mounted high on the side of his boiler, and it works very reliably in this position. I want to move my injectors similarly, because this prevents water from siphoning out the overflow, if a small hole is drilled in the overflow line right below the injector. Also, I wouldn’t have to close the water valve on the tender when shutting off steam to the injector. My only concern is that I may still need to fiddle with the water valve due to varying boiler pressures?

Fiddling with the water valve is the nature of the beast regardless of where you install your injector. The nice thing about having the injector in a lifting position is you will be able to find a water valve setting that works for a wide pressure range, say 80-120psi without adjustment, and you won't have to touch it unless your boiler pressure drops below that range.

On the full sized, the move to non-lifting injectors was really more about maintenance. As injectors got bigger and bigger, having them mounted low became a major advantage. When a lifting injector overheats, the overflow valve can be closed to kick out the warm water from the suction line, and it will then prime. A hot injector with cool feedwater will always work. A hot injector with warm feedwater will never work.

[hoppercar]

I see Anthony mentioned the u.k. injectors. I know a couple people who have these British injectors, that appears to be soldered up from barstock ?........on a good day, they aren't very reliable, and only seem to work in a very narrow pressure range window....(if they work at all)......maybe Anthony can leave us an explanation, of why these injectors are so finicky to get to work ,??

[AnthonyDuarte]

I see Anthony mentioned the u.k. injectors. I know a couple people who have these British injectors, that appears to be soldered up from barstock ?........on a good day, they aren't very reliable, and only seem to work in a very narrow pressure range window....(if they work at all)......maybe Anthony can leave us an explanation, of why these injectors are so finicky to get to work ,??

Where to begin...

The UK Model injector considers the gap between the steam cone and combining cone a "regulating gap." They tightly control this gap so that the injector picks up dry at a specific pressure without having to adjust the water valve. I don't know what they target as their sweet spot, but it seems to be around 80psi, since most of the injectors stop working at around 100psi. The gap does not allow enough water to pass through at higher pressure, so the steam does not condense and thus the injector stops working. When the gap is more than wide enough, you start to see a loss of efficiency and a little bit of dripping at the high end of the pressure range rather than a sharp cutoff of operation.

The problem with the "regulating gap" is that the gap is extremely sensitive. The tiniest changes in diameter or spacing will have a massive impact on the operating range. Even just .0005" of variation will have a noticeable impact on the high end pressure. The steam nozzle is also the most prone to wear. It has steam blasting through the inside, slowly cutting the metal, and it has feedwater water slamming into the outside. Because the steam cone has to be placed so deeply into the combining cone for this "regulating" feature to work, the steam cone ends up with a knife edge. Literally .001-.002 wall thickness right where the nozzle is most prone to wear, so they don't last long. Especially in brass.

A larger gap here has no negative impact on the injector's performance, except for the fact that now the engineer is responsible for regulating the water, rather than leaving it up to the pre-determined gap, which may not even be set correctly depending on variances in manufacturing and assembly.

The UK model design also has significantly wider taper angles than any other injector design, which is fine for low pressures, but it will start to become unreliable at higher pressures. There are, in my opinion, issues with the overflow as well. That signature "chirping" sound you hear is not the sound of a properly working injector, but the sound of one that is constantly sucking in air through the overflow.

[Bill Shields]

I see Anthony mentioned the u.k. injectors. I know a couple people who have these British injectors, that appears to be soldered up from barstock ?........on a good day, they aren't very reliable, and only seem to work in a very narrow pressure range window....(if they work at all)......maybe Anthony can leave us an explanation, of why these injectors are so finicky to get to work ,??

Black magic, prayers and crossed fingers.

[James Powell]

And the blood of at least one failed SAC member

Although we tended to deal with smaller than average injectors (IIRC 7 oz/min on the traction, 11 on the roller & I don't know how small John Crook's were, but they were _tiny_, on a 1 1/4" White Traction...), so there is a bit more logic to the wing on the water and steam and hope for the best approach, as they usually flooded the boiler at about the same time you ran out of something else (fire, steam, hope, cola in your can, cold water...something) which was about 3-5 seconds after starting the whole process.

As such, I kind of am used to the injector being more of a "scenic accessory" than those of you who practice on say, North American 3/4" railway engines bigger than a A5, or anything in 1" or 1.5/1.6".

Axle Pump & (damn it all) hand wagging it was, for the most part. Our 1.5" traction had just the (crankshaft) driven pump and the injector, same as the roller. First check on firing up was always that the pump worked...because I hated to have to hand turn it over >2000 revs in order to refill the boiler if it didn't work and we had to drop the fire !

James

[NP317]

Thinking on injector use, both my 1/8th scale steam locomotives successfully use injectors for boiler water feed from the tender.

So I have been considering installing an injector on my steam launch (people carrying size), until it occurred to me the problem:
Water supply for the condensing steam launch will be from the hot well with temps around 150 F or more. Definitely not conducive to injector use.
So I will complete machining the old Coles Models Duplex water pump that is half done and install it in the launch system. Presumably.

Just thinking out loud to begin 2022.
RussN