Random Thoughts on Steam Injectors

[Anonymous]

I'm still looking at injectors, although in general not bothering to post. However, I have recently bought a reproduction of a book first published in 1893. As expected it contains a fair amount of maths, including calculus. But the interesting thing that struck me is that a lot of drawings of early injectors, and the analysis, use a steam cone that is convergent only. There is no, or only a very short, divergent cone. So that limits the steam at the output to about 0.6 of the input pressure with a maximum velocity equal to the speed of sound in steam at that pressure. Clearly these injectors work, but I've never seen a model injector without a convergent/divergent steam cone.

Since the output pressure of the steam cone is well above atmospheric pressure these injectors cannot be lifting. But that may well not be a problem for a traction engine where the injector tends to be low down on the tender side.

Fascinating stuff.

Andrew

[Michael GilliganParticipant @michaelgilligan61133]

It's good to see that you are persevering with this, Andrew

… Not my area of interest, but I admire your 'need to understand'

MichaelG.

[Anonymous]

Posted by Michael Gilligan on 22/03/2018 16:47:47:

It's good to see that you are persevering with this, Andrew

… Not my area of interest, but I admire your 'need to understand'

That's because I'm an engineer, not a modeller. While it is satisfying to make parts, generally that's the easy bit. What really interests me is how and why the parts are designed the way they are. Plus the drawings are inadequate, and just plain wrong, in some areas. So I am modelling the whole engine in 3D CAD and re-designing along the way as needed, or for interest.

There's no way I could be described as a model engineer, in at least one sense, and probably both senses.

Andrew

[Martin Johnson 1Participant @martinjohnson1]

Hi Andrew,

I have only recently found this thread on your musings about injectors and applaud your inquisitive approach. It might already be known, but the only way you will really understand is to question things and work it out.

Way back you asked about the reason for a particular divergent angle on the steam cone. The answer is to be found in the simple truth:

"Accelerating fluid flow is easy, Decelerating fluid flow is damned difficult."

I have spent a working life in fluid dynamics finding that one out.

More specifically, in a conical diffuser once you reach a certain cone angle you will get flow separation – the boundary layer start to detach at some point downstream of the throat. This is precisely the same effect that causes a wing to stall at too steep an incidence. Initially you will get instability (buffeting) followed by complete BL detachment (stall). Obviously, in a cone of fixed angle you don't have the option of pushing the stick forward to get out of it. What you can do is keep the cone angle below about 10 – 12 degree included angle. Less than this angle and you will get a reasonably controlled (and hence efficient) reduction in flow velocity, turning the velocity energy back into pressure energy. There are charts published of loss coefficients Vs. geometry which show the detail of this effect.

I quite liked your explanation of why a steam nozzle must be convergent – divergent to extend the choking point. You are actually using Bernoulli's theorem to drop the exit static pressure at the throat, hence increasing the p1/p2 ratio which extends the sub-sonic flow regime. Not my specialism, but I hope this provides an alternative way of looking at it.

I have recently been looking at draughting, which is another application of the jet pump (and another area that everybody knows everything about because Kank and Snart said so in 1834). Anyway, one of the more reliable papers indicated that for good cavitation performance the area for the entrained fluid inlet was critical, which translates to radial gap between steam cone and mixing cone, plus axial standoff between same. This makes sense because the lowest pressure area is exactly there. You are trying to entrain water into a jet of steam moving at silly velocity and if you drop the static pressure below vapour pressure, cavitation (and hence loss of performance) will follow. This also explains why lifting injectors or hot water injectors are a *** to design because these areas will be critical.

By the way, I think you need to uplift your design water flow. You have not allowed for cylinder condensation which is likely to put up the water demand by another 50% or so, even allowing for steam jacketed cylinders. See the paper by Bill Hall "Measuring Steam Engine Performance"

Keep up the good work and the analytical approach. I copped out on injector building, there is a unit by Terry Baxter's old firm (He of "Injector Man" fame in another place) which is the vicar's knickers. Works beautifully and is encased in a lost wax casting which is a dead ringer for a Gresham & Craven vertical injector as used on tractions.

Best Wishes,

Martin

[Anonymous]

Martin: Thanks for the detailed explanations. I'm on holiday next week, but the flying weather isn't looking great, so I may well have a lot of time to revisit injectors. Just for completeness I might just finish off this thread by looking at the combining cones and delivery cone.

Your notes on the divergent cone angle are helpful; they've clarified what I sort of understood, but not well enough to expound on it. Regarding aircraft wings in the 1990s there were a number of gliders that didn't exhibit pre-stall buffet. In the UK they couldn't be flown without an external stall warning device. That's a small flap on the leading edge of the wing that flips as the stagnation point moves.

I hope you're planning to write up your notes on draughting and other areas in due course, presumably for ME.

The injector size was based on what TerryB told me. He kindly said I could use his bending rolls for my rear wheel rims. While I was there we naturally talked injectors. He asked about my engine and immediately gave a size. That got me wondering how he knew that, and is what started me on the calculations.

I'm still not sure if I'll ever actually build an injector to my own design. But I remain cautious about commercial items. A friend, who is building the same engine as me, bought one from a well known supplier of boiler fittings. Apart from being the wrong hand and other issues, it looked quite rough, not what I was expecting to see for a premium product. To be fair it was never actually used, so we don't know if it worked.

What I really need to do is retire and then I'd have time to look at this stuff properly.

Andrew

[Cornish JackParticipant @cornishjack]

MJ –"This is precisely the same effect that causes a wing to stall at too steep an incidence. Initially you will get instability (buffeting) followed by complete BL detachment (stall). Obviously, in a cone of fixed angle you don't have the option of pushing the stick forward to get out of it."

I have a feeling that there is a confusion between 'incidence angle' (wing to body) and 'angle of attack' (wing to airflow). Push the stick forward as hard as you like, but you won't alter the incidence angle until you strike terra firma!!

Interesting correlation with flying since modern aircraft fuel systems commonly use 'jet pumps' as tank scavengers and the principles appear to be quite similar

rgds

Bill

[Daniel AcklesParticipant @danielackles42053]

How do you calculate the appropriate sized injector to boiler ratio? I want to build a boiler and purchase an injector from a manufacturer as they seem too difficult for me to build.

I have found an equation: W = (xHv + (Ts – Twd))/Twd – Tws

Here: W = weight of water

x = dryness of steam

Ts = Temperature of steam at absolute pressure

Twd = Temperature of water discharged at injector

Tws = Temperature of water suppled to injector

Hv = Enthalpy of steam at absolute pressure

This has given me some clarity in how to calculate the flow rate of the injector. I understand that the boiler must be appropriate to fit this flow rate. I don't want the boiler to lose pressure or temperature because the flow rate is too fast for the boiler. Is this the correct path to try and understand all of this?

[Jeff DaymanParticipant @jeffdayman43397]

Suggest you find a copy of Mr. D.A.G. Brown's book "miniature injectors inside and out" which has a guide to injector sizing as well as all constructional details needed.

[Speedy Builder5Participant @speedybuilder5]

"find a copy of Mr. D.A.G. Brown's book"

A PDF is also findable on the NET

[Anonymous]

Posted by Daniel Ackles on 13/08/2021 14:00:16:

How do you calculate the appropriate sized injector to boiler ratio?

No idea where your equation came from, but I'm not convinced it's appropriate. The size of the injector is determined by the amount of steam used by the engine. See the first post in this thread for calculations. If the boiler can't provide that steam, or can't accept the calculated injector flow without losing pressure, then the boiler is too small. It is normal to size the injector slightly above the nominal flow rate needed, so it can be used intermittently.

Andrew

[Daniel AcklesParticipant @danielackles42053]

The equation is an approximation of flowrate of the injector. It shows (approximately) the "feed" to the boiler.

I found an injector that I want that claims:

"Hobby Boiler Injector by D & S of Manchester UK. Feed lines are 3/16" and the Steam line is 5/32". Operating range is 75 to 100 Psig and can deliver up to 22 ounces a minute."

If I construct a boiler that is a cube (128 in^3 volume) that is holding roughly 2 lbs of water:

Would I have to ensure that the outlet (supply steam to do work) is greater than the inlet (from injector)? I assume I must… Is there some sort of typical ratio?

[Author]

Posts