|
Post by Roger on Mar 4, 2019 13:19:28 GMT
Would you not expect both injectors to work at full pressure, so if either one fails the other can put water in at a sufficient rate to overcome the safety valves? I am saying this not having looked at any regulations or having yet got to a stage where I need to know, just what I would expect. Baldric. Hi Baldric, That would seem reasonable, unless you have an Axle pump too, whereby it's not a problem. It's interesting that injector failure is almost expected at some point, even though we're told how reliable they are.
|
|
|
Post by delaplume on Mar 4, 2019 13:43:39 GMT
Would you not expect both injectors to work at full pressure, so if either one fails the other can put water in at a sufficient rate to overcome the safety valves? I am saying this not having looked at any regulations or having yet got to a stage where I need to know, just what I would expect. Baldric. Hi Baldric, No, it doesn't work that way---------Remember that the steam being admitted into the injector is already at the Boiler's OPERATING pressure at that moment in time, and it is in the nature of an injector's design that it's output pressure is always above its' input pressure........and it's this pressure difference that overcomes the CLACK valve and thus inject water....NB}---Clack valves have a very lightweight spring, or in some cases non at all, and it is just the boiler pressure that the feedwater pressure has to overcome...........Because the clack valve area is common to both sides this is not needed in the balance equation... As the incoming water--although hot ---- is of a lower temperature than the steam in the boiler, then some of that steam condenses into water, there is a resultant slight transformation of energy for that process to take place, and the boiler pressure starts to drop as a result.... You may have gathered from this thread so far that you can alter and adjust the various cones to get different outputs...The "Holy Grail" if you like is to get an injector that "Picks up" or starts at a low gauge pressure and continues to operate right up to the WP ( stated Working Pressure ) of the boiler under all operating conditions without "Knocking off" or loosing a lot of water via the overflow... The Boiler Inspector is only required to make sure that all feedwater supply systems actually work ... They get tested up to and including the rated WP..... the quantity of injector output is not required to be tested as such.....
|
|
|
Post by delaplume on Mar 4, 2019 13:48:24 GMT
Sorry-----mine's 1st edition--circa 1960.
Yes, I take your point about your preferred method......Have you seen my latest reply that gives a calculator found in a MECH thread many years ago ??
|
|
|
Post by yorkshireman on Mar 4, 2019 13:54:39 GMT
Hallo My Speedy was operated for many years using TWO injectors only, no problems. There is however a 3rd watertank in the bunker behind the cab. This tank was my lifeline, when I struggled to get the injectors going with warm sidetanks on a hot day. The rear tank is not heated by radiation from the boiler. Once the injector had picked up from the cool rear tank, I found by experiment that it was possible to switch the injector feed from the rear tank to a side tank: both lines open to the injector, then closing the 'cold' valve very slowly. In most cases this did work to my amazement. However, I have since added a drivepump.
One could also design a heat-exchanger coil between the sidetanks and the injector, where the warm water from the sidetanks could be cooled in the rear tank.
Spending effort to design one's own injectors is 'love and labours lost', my opinon ... Johannes
|
|
|
Post by Roger on Mar 4, 2019 14:50:46 GMT
Sorry-----mine's 1st edition--circa 1960. Yes, I take your point about your preferred method......Have you seen my latest reply that gives a calculator found in a MECH thread many years ago ?? There's a link to one by John Bauguley, but the link didn't work. There was one mentioned by mutley, that uses the average from someone else's designs. I'm not sure that's a good idea really. I think you're better off using data for the throat sizes and going from there using the ideas in this thread that mostly come from Bob Bramson's book.
|
|
|
Post by Roger on Mar 4, 2019 14:58:39 GMT
Hallo My Speedy was operated for many years using TWO injectors only, no problems. There is however a 3rd watertank in the bunker behind the cab. This tank was my lifeline, when I struggled to get the injectors going with warm sidetanks on a hot day. The rear tank is not heated by radiation from the boiler. Once the injector had picked up from the cool rear tank, I found by experiment that it was possible to switch the injector feed from the rear tank to a side tank: both lines open to the injector, then closing the 'cold' valve very slowly. In most cases this did work to my amazement. However, I have since added a drivepump. One could also design a heat-exchanger coil between the sidetanks and the injector, where the warm water from the sidetanks could be cooled in the rear tank. Spending effort to design one's own injectors is 'love and labours lost', my opinon ... Johannes Hi Johannes, Do you have any insulation between the side tanks and the boiler, other than the boiler lagging? I imagine that most of the heat is going to come from that direction rather than by direct Sunlight. If the rear tank stayed cold in the same weather, that would seem to confirm that conclusion. That's a very interesting observation about the rear tank remaining cold and using it to start the injectors. I'll have the option to fill the side tanks from the driving truck while in motion, and that will be cold too. Hopefully that will help if the tanks become too warm for the injectors. It will also be interesting to see if the second check valve on the overflow really does allow for higher temperature water to be used. I suppose this could be tested on the rig by just removing the second valve and trying it both with and without. I can certainly see why most people wouldn't want to design their own injectors. Sadly, you don't have any option if you want a scale one that works.
|
|
|
Post by yorkshireman on Mar 4, 2019 16:31:30 GMT
Roger The Sidetanks on my Loco do have a constant X-Section, that means that there is a (hidden) air-gap between tank and boiler, widest near the Firebox. I do have a boiler lagging. I did not yet measure the actual Temperature in the tanks, but just from feeling with my hand the sidetanks were notably warmer. At another loco I have the Checkvalve directly after the injector and a second Checkvalve at the boiler feedpoint. Works... Johannes
|
|
|
Post by yorkshireman on Mar 4, 2019 17:03:25 GMT
Sorry I ment to say narrowest at the Firebox... Johannes
|
|
|
Post by Roger on Mar 4, 2019 18:49:39 GMT
Roger The Sidetanks on my Loco do have a constant X-Section, that means that there is a (hidden) air-gap between tank and boiler, widest near the Firebox. I do have a boiler lagging. I did not yet measure the actual Temperature in the tanks, but just from feeling with my hand the sidetanks were notably warmer. At another loco I have the Checkvalve directly after the injector and a second Checkvalve at the boiler feedpoint. Works... Johannes You could still be getting quite a lot of heat from the boiler across the gap by convection. I think it's worth putting a shaped piece of polystyrene in the gap, I think it would make a noticeable difference.
|
|
|
Post by Oily Rag on Mar 4, 2019 19:30:33 GMT
Roger writes, snipped for clarity "In my opinion the book needs to be reworked in conjunction with someone who knows nothing about the subject so these things can be made crystal clear."
This is the profession of a "technical writer" My dear wife's profession that I have learnt a lot about these last 10 years.
|
|
jma1009
Elder Statesman
Posts: 5,896
|
Post by jma1009 on Mar 4, 2019 20:39:14 GMT
Hi Roger,
Do you propose to add a bellmouth 'snout' to the start of the delivery cone? Arthur Grimmett always regarded this as very important and that it should be highly polished.
To link in with Kipford's comments, a 13 degree convergent taper on the steam cone (the angle isn't critical) before the steam cone throat is what Derek Brown, Gordon Chiverton, and Arthur Grimmett and Ted Linden used on their designs, plus Eric Rowbottom and Laurie Lawrence.
The 13 degree divergent taper on the delivery cone on the 'standard' medium size injector 24 - 26 oz per min was simply to sort of fit the injector body length, and IMHO a 9 degree taper would have perfectly acceptable and is what I have used when making replacement delivery cones.
The taper in the condensing cones must be 9 degrees in miniature, except that Eric Rowbottom advised a slight 10 degree peliminary entrance for high pressure injectors when dealing with quite different conditions in South Africa and high altitudes.
As Brian, Chris and others know, I've been sent some very poor commercial UK examples where the combining cone has something like a 12 or 13 degree taper, and they cannot be made to work properly and are 'bin jobs'.
Cheers,
Julian
|
|
|
Post by Roger on Mar 4, 2019 21:36:19 GMT
Hi Roger, Do you propose to add a bellmouth 'snout' to the start of the delivery cone? Arthur Grimmett always regarded this as very important and that it should be highly polished. To link in with Kipford's comments, a 13 degree convergent taper on the steam cone (the angle isn't critical) before the steam cone throat is what Derek Brown, Gordon Chiverton, and Arthur Grimmett and Ted Linden used on their designs, plus Eric Rowbottom and Laurie Lawrence. The 13 degree divergent taper on the delivery cone on the 'standard' medium size injector 24 - 26 oz per min was simply to sort of fit the injector body length, and IMHO a 9 degree taper would have perfectly acceptable and is what I have used when making replacement delivery cones. The taper in the condensing cones must be 9 degrees in miniature, except that Eric Rowbottom advised a slight 10 degree peliminary entrance for high pressure injectors when dealing with quite different conditions in South Africa and high altitudes. As Brian, Chris and others know, I've been sent some very poor commercial UK examples where the combining cone has something like a 12 or 13 degree taper, and they cannot be made to work properly and are 'bin jobs'. Cheers, Julian Hi Julian, Yes, I did mention it in the text, but it's not on the model at the moment. It certainly makes sense when you picture what's happening at that point. I think we should be very cautious in accepting too many things just because a lot of people believe in them. At one time everyone believed the World was flat! The only thing of value is evidence, and Bob Bramson makes all his injectors without a convergent taper or bell mouth at all. That's all you need to know. What's upstream of the throat in the steam cone would appear to be of no consequence. People follow without question because authoritative figures do it, not because there's evidence for it being necessary. From what I can gather, the 9 degree divergent taper is designed to create as near parallel stream as possible. Changing the angle from that probably won't have much impact in reality since the outlet is so close to the Condensing cone. I find it interesting that commercial injectors would be made with angles that are so different from what is known to work. I wonder where their designs come from?
|
|
|
Post by andyhigham on Mar 4, 2019 21:41:28 GMT
I really hope this thread will try to explain why different features of an injector perform in the way they do and add some formulas to aid in design. For too long injector design has been "Make it like this because Mr XYZ says that what works best"
|
|
jma1009
Elder Statesman
Posts: 5,896
|
Post by jma1009 on Mar 4, 2019 22:03:43 GMT
Not so sure about that Roger!
I've never seen or tried a Bob Bramson injector, and neither have you!
I can tell you that the Linden injectors of the 1940s early 50s work extremely well, as do the Arthur Grimmett injectors of the late 50s to circa 1980, as do all the Chiverton Injectors because I have used them and tested them and played about with them. Arthur Grimmett always told me he made over 3,000 injectors for Reeves and Kennions. How many has Bob made?
I have so far not heard of anyone having a Bob Bramsom injector on this forum stating "Yes, I've got one, or I made one, and it it is so wonderful and perfect and utterly reliable"!
(If you were to ask Kipford should the steam cone have a converging taper entrance before the throat as opposed to a simple drilled hole, then we might make things clearer in respect of the steam cone).
Cheers,
Julian
|
|
JonL
Elder Statesman
WWSME (Wiltshire)
Posts: 2,901
|
Post by JonL on Mar 4, 2019 22:08:50 GMT
I can see Roger's point, none of this is magic or witchcraft, theoretically it should be capable of being reduced down to a simple rule or equation. People may have stumbled upon good results in the past through experimentation and common sense, but with proper modelling it could be that a proper technique for production can be come across, especially as nowadays we have the ability to simulate results rapidly through computer simulation.
|
|
|
Post by Roger on Mar 4, 2019 22:38:03 GMT
Not so sure about that Roger! I've never seen or tried a Bob Bramson injector, and neither have you! I can tell you that the Linden injectors of the 1940s early 50s work extremely well, as do the Arthur Grimmett injectors of the late 50s to circa 1980, as do all the Chiverton Injectors because I have used them and tested them and played about with them. Arthur Grimmett always told me he made over 3,000 injectors for Reeves and Kennions. How many has Bob made? I have so far not heard of anyone having a Bob Bramsom injector on this forum stating "Yes, I've got one, or I made one, and it it is so wonderful and perfect and utterly reliable"! (If you were to ask Kipford should the steam cone have a converging taper entrance before the throat as opposed to a simple drilled hole, then we might make things clearer in respect of the steam cone). Cheers, Julian Hi Julian, Indeed I haven't seen a Bramson injector, and I've never made any injector of any type. However, D.A.G. Brown freely admits in his book that he sees no reason to make the Steam cone with any other entry, purely on the basis that it works, so why change it. He's just following along like all the others before and after him without question. Reading Bob Bramson's book, he's clearly no fool, and I'd suggest it's for others to prove him wrong. You only need one injector with a blunt entry to the Steam cone to prove the point! I'm more than happy to try what he says, regardless of how many 'experts' do it a different way, that's how progress is made and knowledge accumulated. This is why there's been so little progress in ME over the past 50 years in my opinion, there's too much deference and not enough cross examination.
|
|
Gary L
Elder Statesman
Posts: 1,208
|
Post by Gary L on Mar 5, 2019 0:50:34 GMT
Could be a case for having one injector made to work at lower pressure and one for full boiler pressure do you think? Then you are covered for the full range of likely pressures. Yes I think so. That is the solution I have adopted anyway. I get the sense that my high pressure injector has a smaller working range than the standard sort, which leaves an inconvenient gap at the low end. It would be nice if the dimensions could be exactly specified for a given max pressure, then you would have the top of the injector's range at your red line, and know that you had the greatest possible range for that particular unit. But that is probably perfection, requiring a degree of accurate manufacture that few of us can aspire to. -Gary
|
|
Gary L
Elder Statesman
Posts: 1,208
|
Post by Gary L on Mar 5, 2019 1:04:12 GMT
...but only at the pressures that LBSC used, which I think is 80psi max, plus say 10psi for luck. The late and lamented Ken Swan wrote (in 1970) that LBSC injectors didn't work on his Bridget at the designed 120psi (as other Bridget/Jessie owners will concur). Ken says in his articles that "after much fiddling about, it now works perfectly even with warm water, but the steam cone is .004 in bigger than the specification" I don't think Ken realised back then that the problem lay with the pressure he was using (he jokes about 'thin water' in Surrey!), but later authors like DAG Brown give a similar recipe for dealing with higher boiler pressures. -Gary Looking at Bob Bramson's book, it would appear that all you have to do is to use a slightly smaller steam cone throat to start with using the same throat diameters for the Mixing and Delivery cones as before. Everything else follows from that, ie the throat of the Condensing cone might need changing slightly to make sure it's still going to work as an Ejector, and that will affect the length of that cone. Bob Bramson gives two columns of figures in Table 2 on page 20 that show the sizes you need for two pressure ranges. Making it work at higher pressure sacrifices it working at quite such a low pressure though, you can't have both. On re-reading DAG Brown, he seems to agree with that, which I'm sure is right but is the exact opposite solution to Ken Swan's. No wonder my brain hurts! -Gary
|
|
|
Post by Oily Rag on Mar 5, 2019 3:05:23 GMT
Roger mentions "I think we should be very cautious in accepting too many things just because a lot of people believe in them." I agree. The plural of anecdote does not = fact.
Model engineering is also about exploring, experimenting, having a go at something. Bravo for this thread, it is very absorbing.
|
|
kipford
Statesman
Building a Don Young 5" Gauge Aspinall Class 27
Posts: 566
|
Post by kipford on Mar 5, 2019 5:52:06 GMT
Rodger What happens upstream of the steam does have an effect. As you know every time you do work on a fluid, expand it, contract it, turn it etc there is a loss. It is the magnitude of the loss what is important. So to answer Julian's question, there is no correct answer. The aim should always be to minimise the losses. The velocity of the steam in the pipe will determine if a plain entry is Ok or whether a larger pipe contracting down to the nozzle throat diameter is better. With jet pumps we always ran a larger pipe contracting down, but we were dealing with relatively high mass flows and wanted minimise the pressure loss in the system to ensure the entrainment ratio (equivalent to the ratio of the mass flow of water from the tender / mass flow of steam through the nozzle) was as high as possible, typically 10:1 or greater. I need to do some sums to understand what values we are dealing with, I have a hunch the injector entrainment ration is a lot less from the relative sizes of the cones, however being on Holiday precludes this at the moment. I have been thinking about the design a lot and when I get home I intend to go off piste and look at a revised design without recourse to what has gone previously. The second thing is I will investigate doing some computational fluid dynamics (CFD) on a conventional design and what ever I dream up. If the CFD works as I hope (our nickname for it at work was painting by numbers), it will be interesting to see what actually go on inside an injector. This will not be the work of 5 minutes, but I will update if it all goes to plan. Going wine tasting tomorrow in Blenheim😋 Dave
|
|