Author Topic: Steam water feed injectors  (Read 27110 times)

Offline sbwhart

  • Full Member
  • ****
  • Posts: 732
  • Live Long and Prosper
Steam water feed injectors
« on: June 16, 2015, 07:38:48 AM »
I've just completed making some steam injectors for my 5" gauge locomotive.

Injectors are a very cleaver little device that uses the flow of steam through a series of cones to pick up water and to feed it into the boiler against the boiler pressure, it's all to do with changes of velocity and pressure I won't even start to explain the physics of it as its way beyond me it enough just to know they work.

This is what they look like.



They are notoriously difficult to make to work as they have critical features that are difficult to actually measure, you have to get to these features in a round about way, and as I found out there are other pit falls just waiting to catch you. I did have a great deal of help from a friend and very talented model engineer (Pete) who gave me a great deal of advice as to the best way to go about making them, he also has a injector boiler test rig, that greatly help.

The first part to make is the body:- this is fabricated from some 3/8" thick brass bar.

First thing is to square the bar up and centre for the 5.5mm through hole postion



Then in the four jaw set up on this centre with a wobble bar.



Turn up and thread 5/16" * 32 ME and ream through 5.5mm



Then drill through for the overflow chamber and clack valve and tap again 5/16" * 32 ME.



Drill through for the water feed and the overflow and make some 5/16" * 32 ME. connectors



Silver solder the connectors and overflow to the body and drill through.



Make and fit a cap for the clack valve



And that's the body made.

Next up I'll show you how to make the taper D bits and how to make and fit the cones.

Cheers

Stew








« Last Edit: June 16, 2015, 07:45:07 AM by sbwhart »
A little bit of clearance never got in the way

Offline jadge

  • Full Member
  • ****
  • Posts: 489
  • Cambridge, UK
Re: Steam water feed injectors
« Reply #1 on: June 16, 2015, 11:51:49 AM »
I've been looking at the theory of steam injectors, with a view to possibly designing one for my traction engines.

The isentropic expansion of the steam in the steam nozzle and the pick up of water by the resultant low pressure, but high speed, steam jet seem fairly straightforward. What I haven't completely got my head round yet is the conversion of the high speed jet back to a sufficiently high pressure to feed into the boiler. How can a flow of steam from the boiler, at a given pressure, inject a larger mass flow of water into the boiler at, or slightly above, the same pressure? I think it has to do with the enthalpy of the incoming steam. When the steam jet is condensed by the feed water this heat energy is given up to the water and it is this extra energy that allows the water to be injected into the boiler. So by this argument an injector wouldn't work using compressed air, as the air would not contain much heat energy.

Andrew

Offline Jo

  • Administrator
  • Full Member
  • *****
  • Posts: 14838
  • Hampshire, england.
Re: Steam water feed injectors
« Reply #2 on: June 16, 2015, 12:18:46 PM »
How can a flow of steam from the boiler, at a given pressure, inject a larger mass flow of water into the boiler at, or slightly above, the same pressure?

Don't forget that the water is cold, when the hot steam draws in and mixes with the cold water it forms a vacumme, which in turn sucks in more cold water the combination of which are at a higher pressure than the original steam, this higher pressure mixture can then force its way into the boiler  ;).

Which is why they don't work with warm water  :ShakeHead:

Jo
Enjoyment is more important than achievement.

Offline Stuart

  • Full Member
  • ****
  • Posts: 1785
  • Tilchestune UK
Re: Steam water feed injectors
« Reply #3 on: June 16, 2015, 12:26:08 PM »
Good work there Stew

Made a few in the past all but one worked , I could not solve it so it went in th bin

A good read on the subject by DAG Brown   
http://www.teepublishing.co.uk/books/operation-valve-gears-injectors/miniature-injectors-inside-and-out/

LBSC though he knew how they worked but a lot of his info was flawed , DAG Brown has it sorted , but you do need a injector rig to test/prove the flow rate

A major fault is faulty soldering ,pin holes and flux inclusions that are ok at first but fail in service

The last set I did was built for my class 4 but built into a scale body with the injector itself pushed in and sealed with o rings very small section ones , sorry I have no pics as they went to a new home but not on the loco when I sold it

Have fun but remember the full sized ones with the crocodile mouth work very well but do not scale , as the size gets smaller they get harder to make at 7 oz a min it's tricky

The other problem is the steam pressure range as you do not always have full steam that's why you need to build a test rig

They do need a clean regularly with very dilute citric acid NO poking the cone with a wire please
And a water filter is a must as is a good clack with a larger lift than used with ram pump
The sound of a injector feeding is music to my ears ( if they still worked)

Got the popcorn out for the long haul

Stuart
My aim is for a accurate part with a good finish

Offline gbritnell

  • Full Member
  • ****
  • Posts: 2394
Re: Steam water feed injectors
« Reply #4 on: June 16, 2015, 12:33:33 PM »
Hi Stu,
Back when I was into steam and built my traction engine I studied the operation and making of injectors. At that time I didn't know enough to even try and with the intricacies of the jets I don't know if I'd still want to try. I'll be following along to see your procedures.
gbritnell
Talent unshared is talent wasted.

Offline jschoenly

  • Full Member
  • ****
  • Posts: 85
Re: Steam water feed injectors
« Reply #5 on: June 16, 2015, 01:46:26 PM »
Wanting to build some small injectors myself, I'll certainly be keeping an eye on the next steps.  Thanks!
Jared
                                 Cabin Fever Expo
--- Model Engineering and Home Shop Hobbies in all Forms! ---

Offline kvom

  • Full Member
  • *****
  • Posts: 2599
Re: Steam water feed injectors
« Reply #6 on: June 16, 2015, 02:13:14 PM »
Interesting project.  On my loco I bought injectors from Superscale, as I doubted I'd succeed making my own.

Offline sbwhart

  • Full Member
  • ****
  • Posts: 732
  • Live Long and Prosper
Re: Steam water feed injectors
« Reply #7 on: June 17, 2015, 07:11:06 AM »
Wow I didn't think this would have so much interest.

I'm afraid I'm going to have to keep you all in suspenders until next week as I'm  a little tied up on domestic duties at the moment , and I didn't photographs making the D bits or the cones at the time so I'll make some spare and walk you guys through the procedure.


How can a flow of steam from the boiler, at a given pressure, inject a larger mass flow of water into the boiler at, or slightly above, the same pressure?

Don't forget that the water is cold, when the hot steam draws in and mixes with the cold water it forms a vacumme, which in turn sucks in more cold water the combination of which are at a higher pressure than the original steam, this higher pressure mixture can then force its way into the boiler  ;).

Which is why they don't work with warm water  :ShakeHead:

Jo

My loco is a tank engine as a result the feed water is warm to the touch, but my injectors work with it, I put this down to my friend Pete advice to have a slightly larger annulus (0.002"/0.05mm) between the steam cone and the combining cone, the theory being is that you need a greater volume of warm water to condense the steam and form the vacuum. I suppose those of you who studied the drawing are wondering how the hell doe's he control the annulus to that level well I'm afraid you'll just have to wait until next week to find out  :disappointed:

 :naughty:

Stew
A little bit of clearance never got in the way

Offline jadge

  • Full Member
  • ****
  • Posts: 489
  • Cambridge, UK
Re: Steam water feed injectors
« Reply #8 on: June 18, 2015, 10:34:55 PM »

Don't forget that the water is cold, when the hot steam draws in and mixes with the cold water it forms a vacumme, which in turn sucks in more cold water the combination of which are at a higher pressure than the original steam, this higher pressure mixture can then force its way into the boiler  ;).

Which is why they don't work with warm water  :ShakeHead:

Jo

I'm not sure I follow that. As far as I can see from my calculations the jet of steam from the steam cone is already below atmospheric pressure. If it wasn't the injector would never pick up in the first place. When the steam and water mix then the steam is condensed, and gives up its heat of evaporation. This heat warms the water. The steam jet is also moving extremely fast, by conservation of momentum the mass of water is accelerated, but not to the same speed as the original steam jet. I think that the output cone then converts this momentum into pressure to feed the boiler. I don't know what happens to the pressure in the combining cone, but I'm not sure why it should rise? I thought that the problem with warmer water was that the steam would not then be properly condensed in the time taken to traverse the combining cones, and operation of the injector falls apart.

Andrew

Offline Jo

  • Administrator
  • Full Member
  • *****
  • Posts: 14838
  • Hampshire, england.
Re: Steam water feed injectors
« Reply #9 on: June 19, 2015, 09:07:24 AM »

I'm not sure I follow that. As far as I can see from my calculations the jet of steam from the steam cone is already below atmospheric pressure.

And it is HOT for Stew's boiler at 90Psi on the gauge, taking into account atmospheric pressure, that would be about 166 degrees C.

Try filling a suitable small glass bottle with water out of a just boiled kettle, then once the glass is hot to the touch pour out the water put your finger/hand over the open neck and put the container in some icy water and you will find that it forms an impressive vacuum without the need for any pressure being involved.

Warmer feed water just forms less of a vacuum so then it is down to how good the design/construction of the injector is.

Jo
Enjoyment is more important than achievement.

Offline sbwhart

  • Full Member
  • ****
  • Posts: 732
  • Live Long and Prosper
Re: Steam water feed injectors
« Reply #10 on: June 19, 2015, 01:12:11 PM »
Thanks for the explanation Jo I guess its the old gas law PV=MRT

I'm afraid its beyond me to give a full explanation about the ins and outs of how these things work.

All I know is that its not easy to make a working one, I'll now try to explain the method I used, I must admit that I had a lot of coaching from friend Pete on this which I would like to thank him for.

So making the cones:-

I'll post this in instalments as the method is rather long winded.

First Job is to make some tooling.

D Bits you will need two tapered D bit whilst you set up make a few spares to allow for breakages.



9 deg and 13 deg made from 3/16" dia silver steel (drill rod) to turn the taper you will need a razor sharp tool set perfectly on centre, to set you compound at the exact angle turn up a test length and run a DTI over a set length and not the deflection of the pointer for the 13 deg over a 10mm length the deflection will be 1.132mm




With the compound set gently turn the bar down until you have a sharp point, and give it a polish.

Part it off and centre drill the back and whilst your at it put a groove into it.



Then on the milling machine using a sharp milling cutter to reduce deflection of the point setion the D bit so that it is 0.003" 0.1mm above the centre line.



Use the groove to hang the D bit from a bit of wire , this is so you have something to hold onto, heat to cherry red and quench vertically into water, so that it stays straight.



When you make the 9 deg bit with the compound set make the assembly tool for the combination cones, this has to be made as accurately as you can.



To get the 0.101"(2.565) run out correct first turn a parallel diameter to this size, then with the compound gently turn the point down until it just runs out at the shoulder this shoule give you the correct size. then accurately turn the 0.431" (10.95mm) length.

You will also need to make a 0.037" (0.94mm) thick washer and a setting ring drilled with a No 38 drill, and some stop sleeves for the D bits

The cones have to be made in the correct sequence as you use one to make the other.

Start first with the combination cones:- turn the dia down to be a push fit into the body and part off two slugs one 0.162" (4.11mm) long the other 0.213" (5.4mm) long.



These little slugs are not easy to hold true, I solved this by using soft jaws bored out to a step to grip the slugs



You will also have to set depth a stop on the 9 deg D bit using the setting ring.



Start with the shorter of the two slugs 0.162" check that it is running true, and centre drill and drill through No 56 then its best that use use the D bit floating so with ut held in a carrier and pushing with the tail stop cut the taper , position the d Bit flat side is underneath or on its side if you have it on top the swarf will build up and probably break it.





Then using a triangular scraper radius the mouth of the cone.



Remove the cone from the chuck, now your going to use it along with the washer to cut the next cone



Put the thicker slug in the chuck checj that its running true centre drill and drill No56 and using the D bit with the first cone and washer to cut the taper to the correct depth, with the triangular scraper radius the mouth of the cone, its important you do this things won't work without it.



With the tapers correctly cut in the cones all that required now is to finish off the stepped diameters





To assembly the cones into the body you know use the assembly tool to get them to the correct depth check that you have a gap between the two cones of 0.037" (0.94mm) by using a No 63 drill as a gauge through the over flow chamber. As a belt and brasses action secure with a tiny drop of High temp loctite 640.



 

Phew

That took some explaining.

Stew

The Steam cone is next but I'm going for a brew first.






A little bit of clearance never got in the way

Offline sbwhart

  • Full Member
  • ****
  • Posts: 732
  • Live Long and Prosper
Re: Steam water feed injectors
« Reply #11 on: June 19, 2015, 03:25:47 PM »
OK on with the Steam Cone



This cone and its location is the most critical part of the injector so its worth taking your time and if you have any doubts start again.

First job is to turn up the blank make it a nice slide fit in the body for a length of 1/2" (13mm) there are a number of pits falls her, so slightly undercut the shoulder so that it will but up flat against the body and won't be held off by any radius in the corner, also with a a radius tool just skim a short shoulder this is to avoid a false assembly from any residue of adhesive from the combining cone.



Face the length down so that the gap between the body is 0.094" measured with a No 63 drill, and set the saddle dial at zero, this is important.





Set the 9 deg D bit depth with a setting ring drilled with a No 40 drill, and centre drill, drill No 56  and cut the cone.
This D bit operation will swell the diameter so its best to do this before the next operation, I fell into this trap.



Now I will say this only once this is important. taking small cuts .001" reduce the diameter of the end for a short length with the radius tool until the body just buts up flat with the cone, then face off by a further 0.0015" (0.38) if you set you're dial at zero this is quite easy. This gives the steam cone the correct insertion into the combination cone with the correct annulus.





You should have a cone with an end that looks like this.



Part off

Set back up in the collet face the head off to length  centre drill.



13 deg D bit  depth set with a 1/8" diameter setting ring.



And finally radius the mouth with a triangular scraper.



Only one part left now.

Stew



A little bit of clearance never got in the way

Offline sbwhart

  • Full Member
  • ****
  • Posts: 732
  • Live Long and Prosper
Re: Steam water feed injectors
« Reply #12 on: June 19, 2015, 03:50:08 PM »
Delivery Cone

Turn up the blank again making it a nice slide fit in the body, and face the length off so that it just buts up against the body, zero the dial and reduce the length by 0.024" (0.61mm) this form the gap for the over flow.



With the radius tool turn the end down for a short length 3.5mm to a diameter of 3.2mm, just touch the face with the centre drill to give the No 68 drill a start and drill about 8mm deep, keep drawing the drill back and clean away any build up of swarf from the flutes or you'll end up breaking it.





And radius the mouth with the triangular scraper this is important.



Part off

Face the collar off to length centre drill and 13 deg D bit and radius the edge with the scraper



That's it with a bit of luck you should have a Injector that works.



These are the tools.



Injectors are a bit of a challenge to make you have to follow the procedure carefully and there are lots of little pitfalls that can catch you out, so you have to work very methodically, I got paranoid about getting the drills mixed up so I threaded them through a bit of paper.



And I found it paid to keep the D bits sharp on an oil stone.



But its a great sence of achievement when you get them working, and it gives you the bragging right at the track to say of course I make my own.

 :Jester: :Jester: :Jester: :Jester:

Stew

I'll try and take a Video of the injector working the next time I steam my loco.

« Last Edit: June 20, 2015, 08:33:06 AM by sbwhart »
A little bit of clearance never got in the way

Offline BillTodd

  • Full Member
  • ****
  • Posts: 491
  • Colchester UK (where the lathes were made)
    • Bill's website
Re: Steam water feed injectors
« Reply #13 on: June 19, 2015, 05:35:36 PM »
Nice job StewartB-)

Thanks for explanation Jo , I saw someone demo'ing one on youtube vthe other day and was wondering how the thing could work! :-)

Offline Jo

  • Administrator
  • Full Member
  • *****
  • Posts: 14838
  • Hampshire, england.
Re: Steam water feed injectors
« Reply #14 on: June 19, 2015, 05:43:24 PM »
8) looking forward to seeing it work Stew.

I recommend that you make a batch before you loose those reamers or they have their edges damaged in the drawer  ;)

Jo
Enjoyment is more important than achievement.

Offline mike mott

  • Full Member
  • ****
  • Posts: 418
  • Alberta Canada
Re: Steam water feed injectors
« Reply #15 on: June 19, 2015, 07:09:25 PM »
Thanks for the detailed step by step Stew, I have filed this away for when I am confident enough to have a go at making one down the road.

Mike
If you can imagine it you can build it

Offline Steamer5

  • Global Moderator
  • Full Member
  • *****
  • Posts: 1249
  • The "Naki" New Zealand
Re: Steam water feed injectors
« Reply #16 on: June 20, 2015, 05:10:32 AM »
Hi Stew,
Only  :popcorn: harmed during the following of this thread! Oh an the occasional :DrinkPint:
Nice "how to" article.

BTW I recently came across what I think is a one of the best article on injectors.....I've been collecting articles on these little beasties for some time......the article is written by Bob Bramson it runs to some 26 pages and will answer most if not all the questions that have been asked. For those that may be interested check out The Whistle magazine for the British Columbia Society of ME.... www.Bcsme.org starts in the April 2012 issue, September, October January 2013, February March April May, to save you hunting around.
Bob has also written a very good article on braking for those of us into locomotives, which is how I found the above, if anybody has a design for a vacuum brake valve I would love to here from you!

Cheers Kerrin
Get excited and make something!

Offline jadge

  • Full Member
  • ****
  • Posts: 489
  • Cambridge, UK
Re: Steam water feed injectors
« Reply #17 on: June 20, 2015, 10:34:05 AM »
Kerrin: Thanks for the link; at first glance the articles look to be a good mix of theory and practical. In due course I'll sit down and read through them in detail. I don't envisage a problem in building an injector, designing one is a whole different ball game, but I'm keen to have a go, at least on paper.

Stew: I'm afraid that steam doesn't generally follow the ideal gas law. Steam is normally considered to be a vapour. Having said that experimentally the expansion of steam in a cylinder is generally taken to be close to constant PV, aka hyperbolic expansion.

Andrew

Offline jadge

  • Full Member
  • ****
  • Posts: 489
  • Cambridge, UK
Re: Steam water feed injectors
« Reply #18 on: June 21, 2015, 09:44:04 AM »

I'm not sure I follow that. As far as I can see from my calculations the jet of steam from the steam cone is already below atmospheric pressure.

And it is HOT for Stew's boiler at 90Psi on the gauge, taking into account atmospheric pressure, that would be about 166 degrees C.


I agree that dry saturated steam at 90pisg will be around 166°C. However, I do not think that the steam at the output of the steam will be at that temperature. If it was, and given that the pressure is at, or below, atmospheric that would imply the steam had a significant degree of superheat. I think that the expansion of steam through a nozzle is taken as being isentropic, ie, the entropy doesn't change. If we assume that the output steam is at 10psia, then the temperature will be at about 90°C. That implies that the enthalpy has changed. Change of entropy is defined as change of energy divided by the average temperature. So if the heat energy in the steam has changed, but the overall energy in the steam hasn't changed, where is the missing energy? It's been converted into kinetic energy, as the steam issuing from the nozzle is moving much faster than the input steam.

From a practical point of view the expansion is not perfectly isentropic as some heat is lost to friction and to radiation from the nozzle body. I wonder if there is any advantage in making an injector from an insulating material, or at least the cones?

Andrew

Offline steamer

  • Global Moderator
  • Full Member
  • *****
  • Posts: 12008
  • Central Massachusetts, USA
Re: Steam water feed injectors
« Reply #19 on: June 21, 2015, 04:12:54 PM »
Hi Andrew!

Interesting points!    I do know that a hot injector is very unlikely to function, and will often not work at all!

This gets worse in the small scale ones, and you have to turn the steam on quickly so as to pick up water and flood them, and then back off on the steam till the over flow stops....

Don't know about an insulating material....something to think about though....

About as close to perpetual motion as it gets these things.    98% efficient as a feedwater heater....about 2% efficient as a pump.

Dave
"Mister M'Andrew, don't you think steam spoils romance at sea?"
Damned ijjit!

Offline robmort

  • Jr. member
  • **
  • Posts: 3
Re: Steam water feed injectors
« Reply #20 on: August 09, 2015, 07:26:41 PM »
Stew,

thanks for the detailed and very useful blog.

A couple of questions:
1) You seemed to be working from a book or printed instruction for the design; is it yours or which other, and is it available?
2) the delivery cone is usually recommended to have a 6deg  cone angle, but you seem to use 9deg. Why is that?

Rob

 

Offline sbwhart

  • Full Member
  • ****
  • Posts: 732
  • Live Long and Prosper
Re: Steam water feed injectors
« Reply #21 on: August 10, 2015, 07:16:32 AM »
Hi Rob

Thanks for your interest to answer your ?


A couple of questions:
1) You seemed to be working from a book or printed instruction for the design; is it yours or which other, and is it available?
2) the delivery cone is usually recommended to have a 6deg  cone angle, but you seem to use 9deg. Why is that?

 

The instructions I'm working from are a step by step guide provided by a friend of mine, if you wish I can scan them and send you a copy just send me a PM with your email address.

From Martin Evans Book " Manual of Model Steam Locomotive Construction" he does indeed have a 6deg delivery cone, The design I work to was derived by my friend after much trial and error he did say that he found the cone angles were not very critical for the injector to work correctly, the critical part is positioning the cones and the only reason for making accurate reamers is that it aids the positioning of the cones easier.

Hope this helps

Stew
A little bit of clearance never got in the way

Offline robmort

  • Jr. member
  • **
  • Posts: 3
Re: Steam water feed injectors
« Reply #22 on: August 10, 2015, 10:03:27 AM »
Stew,

thanks, i've sent a PM.

Regarding the delivery cone angle, it will be interesting to see how well it works at 9° so hope you report the results.
In fact several sources recommend 6° e.g. DAG Brown, and the experiments reported at http://www.modeng.johnbaguley.info/injectors/injectors4.htm ".............I made a new delivery cone with a 6° angle. This was a big improvement......Obviously the problem was due to the incorrect angle".

Rob





Offline sbwhart

  • Full Member
  • ****
  • Posts: 732
  • Live Long and Prosper
Re: Steam water feed injectors
« Reply #23 on: August 10, 2015, 10:21:56 AM »
Hi Rob

I made a batch of 5 and they all work fine


Stew
A little bit of clearance never got in the way

Offline jadge

  • Full Member
  • ****
  • Posts: 489
  • Cambridge, UK
Re: Steam water feed injectors
« Reply #24 on: August 10, 2015, 12:06:35 PM »
I agree with Stew, I don't think that the actual cone angle is particularly critical. From what I have read regarding full size injectors the most important dimension for the delivery tube is the minimum diameter, as this determines the mass flow of water. This of course is the important number. The delivery tube needs to generate a pressure greater than boiler pressure, but if the mass flow is too small having the correct pressure is not much use.

Full size design notes indicate that the length of the delivery tube is important, as well as it needing to be divergent. If the angle is too steep, with a short tube, the flow of water breaks away into turbulent flow, with resultant reduced pressure. The design notes I have use a curved delivery tube, ie, not a cone, but in fact parabolic, I think? Whether this is of any importance in a small scale injector is a moot point. If I ever get as far as designing my own injector I will try a curved delivery tube. I would make a template, using the CNC mill, and then use the template with the hydraulic copy unit on my lathe to make a curved 'reamer'. Then making the delivery tube is as per standard.

I think that one of the problems with warm feed water is the time taken to condense the steam and hence the length of the combining cones. In full size experiments were done with movable combining cones to widen the range over which the injector would operate. I don't know whether this system was ever used on 'production' injectors.

Andrew

Offline Greg t

  • Jr. member
  • **
  • Posts: 1
Re: Steam water feed injectors
« Reply #25 on: September 02, 2017, 06:53:02 AM »
Hi I'm new here,I am looking at making my own injector,your step by step discription is awesome,I have one question to start with,how many oz per min does your injector inject,I'm looking at making one to do about 40oz per min what dimensions will I have to change to achieve this?
It's better to try and fail than not to try at all,

ChuckKey

  • Guest
Re: Steam water feed injectors
« Reply #26 on: September 02, 2017, 09:39:31 AM »
There are several good books on model injector design and construction, incuding 'The Model Injector' by Ted Crawford, published by Australian Model Engineering, and 'Miniature Injectors Inside and Out' by DAG Brown, published by TEE, which I would recommend. For the full-size history, 'Practice and Theory of the Injector' by Strickland Kneass, published in 1894, and reprinted by Lindsay Publications is also very interesting.   

Offline Jon Steensen

  • Jr. member
  • **
  • Posts: 1
Re: Steam water feed injectors
« Reply #27 on: August 26, 2021, 07:41:44 AM »
At first a steam injector seem a little counter intuitive, and in fact it took me some time to understand what is going on.  Where does the energy to pump in the water come from? It seems logical that it should take the same amount of energy to pump the steam back in the boiler as you will gain by letting it out. So you’d think there is no energy left to force the water against the pressure inside the boiler. However this is obviously not true, as an injector has been demonstrated to work many times.

I could go all technical and talk about enthalpy, entropy and adiabatic processes, but to not lose those why are not firmly rooted in thermodynamics, I will try to give a more intuitive explanation.  The thermodynamic stuff is for engineers like me, who would do the design of the injector, and it is a bit complicated, or at least requires a lot of calculations.

The key is that the steam condenses before it re-enters the boiler. The amount of work that is required to force a fluid against a pressure depends on two things: How big is the pressure and how big a volume do you need to force in. This is quite intuitive for everyone who has tried to inflate a bike tyre with hand pump. The higher the pressure in the tire, the harder you have to press the pump handle, and it takes more energy to pump in ten strokes than it does to do one. The key here is that a certain amount of steam has a much larger volume than it did once it was pumped into the boiler as water, and hence you can extract much more energy when letting the steam out than you spend to force it in as water (the 10 strokes out, one stroke in thing). This also explains why a steam engine can drive its own feed pump.
If you imagine a piston with water on one side and steam on the other, the pressure on the steam side is just sufficient to press in the water against the pressure of the same side. However by doing so we put in way more water than is required to compensate for the steam we let out, and the rest we can use to do some other useful work.
The same is true for an injector, where the volume of steam we take out the boiler is way higher than the amount of water it will be turned into, when it is returned into the boiler. But in order to understand how exactly it is possible we need to understand another fact.

Imagine drilling two holes in a pressurised boiler, one above the water line and one below it and observe what happens. You will of course see a stream of water being expelled at one of the holes and a jet of steam coming from the other, but what is more important, you will see that the steam is moving really fast, whereas the water is going a lot slower. This is due to the fact that the steam is much lighter than the water, and hence the force due to the pressure drop can accelerate it to a greater speed. It is like you can make a bike accelerate a lot faster than you can make a car go when you push it, since the bike is so much lighter and hence easier to accelerate
The pressure drops when the fluids exit the boiler and the fluids speeds up, as pressure is converted to speed. This process can also be run in reverse: make a speeding fluid stop, and the pressure will rise. This is why you feel force from a jet of water press against your hand when it hits it and you hand slows the water down, and this is exactly what happens in the delivery cone. Since water is so much heavier than air, it will reach a greater pressure when slowed down. You will experience that blocking a stream of water going only a few meters per second generates a significant force, whereas you can hardly feel a gust of wind at the same speed.
It is just like the flow of water and steam (a light gas like the air) flowing out of the boiler, but in reverse.

In an injector steam is taken out, and accelerated to great speeds in the nozzle, whereafter it is combined with a stream of cold water. In the process the fast steam slams into the water, which will slow down the steam and speed up the water until both streams travels at the same speed.
Since the steam loses speed to the water, the mixture cannot speed up to the same speed as the steam alone had, as its energy now has to drive both the steam and the water. But here comes the important part: It does not have to for it to presss the mixture back into the boiler.
This is due to the fact that once the steam mixes with the water it gets cooled down so it becomes water; and as we experienced with our punctured boiler and the jet of water hitting our hand, water will give a much higher rise in pressure when you slow it down than steam will, and therefore it can run up against the pressure of the boiler going a lot slower than the steam did.

Since the “trick” in an injector is that the steam condenses when meeting the cold water, we also find the explanation for why a stream of warm feed water is problematic  there, and why you need to bypass the injector to get rid of warm water that may be stuck in there. If the feed water is too warm, it cannot absorb enough energy from the steam without the water being heated to its boiling point and becoming steam, which is much harder to pump into the boiler.
An injector cannot be driven on compressed air either, since the air will not reduce in volume by condensing like the steam and hence it requires almost the same amount of energy to pump back into the boiler as it gains by going out of it, leaving no energy to drive in the water.

Since an injector depends on the condensation of the steam, designing an injector is a tricky balancing act. Make the steam nozzle too small and you do not have enough steam to speed up the water to sufficiently  to cause  a big enough pressure rise so it can enter the boiler. Make the nozzle too big and you add too much warm steam to dump the heat energy in the water, and you cannot make it condense and become easy to get back into the boiler.
The combining cone has to be long and narrow in order to allow the fastflowing mixture to stay in it for long enough to give the steam enough time to condense, and the delivery cone has to be long and slender to give a nice gradual convertion of speed to pressure in order to not have step pressure gradient that can cause turbulence. Turbulence will cause an energy loss, and if becomes too bad, too much speed is lost and the pressure cannot be raised sufficiently to get the water in the boiler.

Offline Admiral_dk

  • Full Member
  • *****
  • Posts: 3016
  • Søften - Denmark
Re: Steam water feed injectors
« Reply #28 on: August 26, 2021, 11:28:28 AM »
Thank you for the explanation   :praise2:

While it seems a bit long during the read - it all makes good sense @ the end - because of the whole explanation ...!... so if any of you are interested in the why - please read it all.

Best wishes

Per

Online Kim

  • Global Moderator
  • Full Member
  • *****
  • Posts: 5915
  • Portland, Oregon, USA
Re: Steam water feed injectors
« Reply #29 on: August 26, 2021, 06:35:13 PM »
That is a great explanation!  It's the first time I've had a sense of how/why injectors work.

Thanks for taking the time to write that explanation for us, Jon,
Kim

Offline Chipmaster

  • Full Member
  • ****
  • Posts: 918
Re: Steam water feed injectors
« Reply #30 on: August 26, 2021, 08:52:57 PM »
Thanks Jon

Andy

Offline RReid

  • Full Member
  • ****
  • Posts: 985
  • Northern California
Re: Steam water feed injectors
« Reply #31 on: August 27, 2021, 01:18:26 AM »
But if you've ever operated an early, two valve injector, and heard with your ears and felt in your hands the water and steam mixing and flowing and forcing open the check valve, you know the truth. They work by magic.
Regards,
Ron

Offline AVTUR

  • Full Member
  • ****
  • Posts: 210
  • West of England
Re: Steam water feed injectors
« Reply #32 on: August 27, 2021, 12:17:14 PM »
Jon

Thanks for the explanation on what a complex subject. This led me to look at your profile and I see that it is your first posting.

Trying to be very polite, please could you introduce yourself.

AVTUR
There is no such thing as a stupid question.