|
|
Post by John Baguley on Feb 11, 2021 0:04:14 GMT
I've just come across this article in ME for May 1900. It describes a model injector made by a Mr. Henry Lea.
It looks like he beat you to the idea of the one piece combination and delivery cone Roger! Apparently, he found that it worked better than separate cones. As the saying goes, there's nothing new under the Sun. He says that it had a 'bore' of 0.018" but not sure what that refers to. It doesn't appear to have any ball valve on the overflow and has an adjustable steam cone.
Reading further, it appears that the injector worked from 20psi (must have been a very low pressure boiler) down to 1 or 2psi. He does say that it wouldn't start at first and he had to add a pet cock in the delivery pipe which was opened until the injector picked up when it could be shut.
John
|
|
|
|
Post by Roger on Feb 11, 2021 9:17:51 GMT
Does anyone know who designed the original model Injector arrangement that we all know so well? I feel sure it was inspired by the sort of design in the picture. I guess they discovered that the Overflow Valve wasn't strictly necessary for our purposes, so they left it off. Personally I think that was a mistake, you don't want air to be drawn into the boiler, it compromises the performance when the Water temperature is high. There's an article in Model Engineer for September 1898 describing the 'Vic' type injector being sold by a W. Eaton which is the common type we know today. I suspect the actual design goes back before then.
John
Hi John, Thanks for that, I didn't realise it went as far back as that. It's interesting that the only development seems to have been to add a second valve for the Delivery overflow. Surely I can't have been the only person to see that one common one ought to be enough? |
|
|
|
Post by Roger on Feb 11, 2021 9:43:06 GMT
I've just come across this article in ME for May 1900. It describes a model injector made by a Mr. Henry Lea.
It looks like he beat you to the idea of the one piece combination and delivery cone Roger! Apparently, he found that it worked better than separate cones. As the saying goes, there's nothing new under the Sun. He says that it had a 'bore' of 0.018" but not sure what that refers to. It doesn't appear to have any ball valve on the overflow and has an adjustable steam cone.
Reading further, it appears that the injector worked from 20psi (must have been a very low pressure boiler) down to 1 or 2psi. He does say that it wouldn't start at first and he had to add a pet cock in the delivery pipe which was opened until the injector picked up when it could be shut.
John
Hi John, I was just thinking that would be difficult to start, and then you mention it. Maybe at lower pressures it was just about manageable. As you say, there's not much new. It takes a lot of experiment to find an arrangement of holes that work well enough for the Combining Cone overflow, so maybe that's why the separate cones have ruled supreme... until now. |
|
|
|
Post by delaplume on Feb 11, 2021 11:56:59 GMT
Frenchman Msr.Henri Giffard is generally deemed to be the inventor of the full size live steam injector......Have a look at this}---- en.wikipedia.org/wiki/Injector. Air is of course an oxidising agent ( even more so at elevated temps. that the typical boiler operates at ) and thus definitely a NO-NO in steel boilers.. |
|
|
|
Post by Roger on Feb 11, 2021 23:37:44 GMT
I completed another batch of Injector blanks this morning and have made another three cones for experiments. The good results continue as I've been exploring ever larger holes and trying different positions for them. 1mm holes seems to be a step too far, but 0.85-0.95mm seem to work the best. I can get all of these to be dry from 90-40psi and down to about 25 with a bit of care. However, to get that bottom end performance, the regulation gap has to be as small as possible, consistent with just remaining dry at 90psi. Closing the gap too much causes the delivery pressure to drop dramatically so the sweet spot seems to be when you are just at the point where any closer starts that rapid drop in outlet pressure and the overflow starts to weep.  20210211_232154 by Georgia Montgomery, on Flickr I think the 0.8mm radius I was adding may have been a bit too big, so I've made a new radius tool from 4mm Silver Steel which is 0.4mm  20210211_224409 by Georgia Montgomery, on Flickr  20210211_225745 by Georgia Montgomery, on Flickr As you can see, there's a huge difference when they're held side by side. The result is that it's noticeably quieter throughout the whole pressure range and it doesn't appear to have affected the low pressure performance. I may have a look at the outside of the Steam Cone and add a more generous radius there. I'm still not entirely happy with the picking up at the top pressure, I could still use a bit more suction. I'll probably add this small radius to the other more promising cones to see if any of those show bigger improvements than this one. There's still a bit more experimenting I can do with the position of the overflow, and if it's sufficiently upstream, I might even be able to use a second row of small holes to help with starting. However, that might prove to be detrimental to the low pressure performance. All of these tests have been performed with water close to 20C and with the 1mm Steam Cone. I still have the 1.05mm Steam Cone to experiment with.  20210211_230107 by Georgia Montgomery, on Flickr |
|
|
|
Post by chris vine on Feb 12, 2021 12:16:06 GMT
Hi Roger,
You are making grand progress.
I am thinking about the flow of the water past the nose of your steam cone and Julian's comments about annular regulation etc.
If you find that your radius on the front of the cone above helps to make a more stable flow (and better range), I wonder if you can try a small, (concave) radius on the nose of the steam cone?
Clearly this will give a different flow for an end-gap setting, but it might be a compromise between your end gap and the annular gap. I can see that the water is being guided into the steam jet in the normal annular arrangement, but the fact that you have got so far with your system means that maybe it doesn't matter too much.
Equally, you have probably experimented more than anyone else with the overflow arrangements, so your cones might in fact be much better! Maybe a half-way-house of a convex radius on the nose of the steam cone would help...
Chris.
|
|
|
|
Post by Roger on Feb 12, 2021 12:48:20 GMT
Hi Roger, You are making grand progress. I am thinking about the flow of the water past the nose of your steam cone and Julian's comments about annular regulation etc. If you find that your radius on the front of the cone above helps to make a more stable flow (and better range), I wonder if you can try a small, (concave) radius on the nose of the steam cone? Clearly this will give a different flow for an end-gap setting, but it might be a compromise between your end gap and the annular gap. I can see that the water is being guided into the steam jet in the normal annular arrangement, but the fact that you have got so far with your system means that maybe it doesn't matter too much. Equally, you have probably experimented more than anyone else with the overflow arrangements, so your cones might in fact be much better! Maybe a half-way-house of a convex radius on the nose of the steam cone would help... Chris. Hi Chris, I agree that there's room for experiment with the nose of the Steam Cone. I think you mean Covex though, not Concave. Bob Bramson's book shows illustrations of what he calls End Regulation, but in fact only one of those truly has a portion of the gap that's truly perpendicular and set by the Axial position of the Cones. The other clearly shows that the nose diameter of the Steam Cone is smaller than the inlet diameter of the Condensing cone. Clearly the gap is then decided by the closest proximity of the two cones. I think the takeaway here is that there's a solution that's neither strictly End or Annular regulation, but somewhere in between. Getting the flow so it's not absolutely perpendicular to the axis certainly makes sense. The only problem with going down that road is that it makes it pretty much impossible to just make a set of cones with fixed dimensions and have the regulation gap just right. Tiny changes in the shape of the nose of the Steam Cone or the radius of the Condensing Cone will alter the gap, and those are hard to control. That's not an issue for me, I'm going to make the Steam Cone adjustable, even though that ought not to be necessary. Having plated with the adjustable one on the development Injector, I wouldn't want to go back to messing about with shims and spacers, it's too painful. |
|
|
|
Post by chris vine on Feb 12, 2021 12:56:39 GMT
Hi Roger,
I was meaning concave in the sense that it would match the shape of the radius in the next cone. IE to start to guide the water in the correct direction.
However a convex (outside radius (like breaking the sharp edge)) might be same/better/worse! I think once you have something approaching what you desire, then these tiny tuning tweaks will be quite easy to see their effects.
Chris
|
|
|
|
Post by Roger on Feb 12, 2021 13:57:24 GMT
Hi Roger, I was meaning concave in the sense that it would match the shape of the radius in the next cone. IE to start to guide the water in the correct direction. However a convex (outside radius (like breaking the sharp edge)) might be same/better/worse! I think once you have something approaching what you desire, then these tiny tuning tweaks will be quite easy to see their effects. Chris Ah, I see. Yes, that's the sort of thing I was thinking about. This is why I've left this sort of thing until now because they can mask deficiencies in the fundamental design. They certainly make a difference. |
|
|
|
Post by andyhigham on Feb 12, 2021 14:39:55 GMT
Please correct me if I am wrong.
A larger regulation gap works better at higher pressures, while a smaller gap is required for lower pressure.
A development could be to have an adjustment lever in the cab
|
|
|
|
Post by Roger on Feb 12, 2021 15:48:34 GMT
Please correct me if I am wrong. A larger regulation gap works better at higher pressures, while a smaller gap is required for lower pressure. A development could be to have an adjustment lever in the cab This is true for a single injector that needs to work over a wide range of pressures. You have to set the regulation gap for the maximum pressure, and there's a bit or wriggle room in that. It will work at high pressure with the gap on the small or large side of what is clearly a sweet spot, where the flow is the smoothest and the overflow is dry. However, if you set it to that position, you'll find that it drops out quicker as the pressure drops because the gap is larger than it needs to be. It's all about what you consider to be acceptable performance. So the answer to your question is yes, you could arrange for the gap to be adjusted by a lever in the cab for lower pressure, but in reality, the effect is pretty much the same as restricting the water with a lever in the cab. There's nothing to stop you setting up the Injectors with slightly different gaps to optimise the performance for higher and lower pressures with precisely the same cones. I may well do this if it means that the injector picks up better if I optimise one for 90psi and the other for say 80psi. Both would work at 90psi and down to say 40psi, but the lower one would be easier to coax down to say 20-25psi It's remarkable that these work dry over such a wide pressure range without any adjustment. |
|
|
|
Post by delaplume on Feb 12, 2021 16:05:45 GMT
Isn't that what moving cone injectors are all about ??.........ie to automatically adjust to changing conditions ...
|
|
|
|
Post by Roger on Feb 12, 2021 16:14:01 GMT
Isn't that what moving cone injectors are all about ??.........ie to automatically adjust to changing conditions ... I have no idea Alan, I imagined it was just about starting them. I guess you could contrive a setup where the gap automatically adjusted itself to be smaller as the pressure drops. It's something I have contemplated, but there isn't much room to fit anything. I wondered about a sprung sliding sleeve that's forced back when the pressure is high. The trouble with any moving parts is the issue of dissolved solids messing with the fits and clearances in our scale. I'll leave that one for someone else to experiment with. The bottom line is that an injector that works from 90-40psi without adjustment is probably plenty good enough for what we need in my opinion. If it goes a lot lower with adjustment, that's a bonus. |
|
|
|
Post by delaplume on Feb 12, 2021 20:44:51 GMT
Yes--- that sounds about right to me...Maybe add on "reliability" with "affordability" and you've got it cracked, I'd say...
|
|
|
|
Post by coniston on Feb 12, 2021 20:52:33 GMT
Isn't that what moving cone injectors are all about ??.........ie to automatically adjust to changing conditions ... The bottom line is that an injector that works from 90-40psi without adjustment is probably plenty good enough for what we need in my opinion. If it goes a lot lower with adjustment, that's a bonus. From my experience of driving locos for many years I whole heartedly agree with you Roger, these will be the pressure range you would expect to see in 'normal' operation and that is when a reliable injector is so helpful. Just knowing it is working without recourse to leaning over to see the overflow is such a bonus. Well done with your experiments, it is very interesting. Chris D |
|
|
|
Post by Roger on Feb 12, 2021 21:21:16 GMT
This is Cone 51, I think I could make these in my sleep! So far I think this is the most promising. I've gradually migrated the Combining Cone overflow upstream to where it's almost half a miilimeter from where I'd originally calculated it to be. In fairness, the one at the exact position would probably have worked fine without the back pressure of the sprung overflow valve, but that has to be taken into account. All along I've struggled to get enough suction to overcome the back pressure, especially when the body is roasting hot. Although there's a small amount of leakage past the valve that helps with cooling, I can't really rely on that. What I need is enough suction so that whatever the situation, the overflow opens and water is drawn through and out of the overflow. Without that, it's going to be unreliable and a nuisance. So the current design uses 8 x 0.9mm holes drilled 4 at 90 degree intervals then the remaining 4 offset by 45 degrees to fill in the gaps. That seems to work really well. I broke another cone earlier being heavy handed, and the holes look spot on. At the moment I'm getting the most reliable pick up at 90psi and that goes down to 40psi dry. Then it can be coaxed down to below 30psi with a little regulation. I've got a few more tests to do, and then I'll post some footage to show some of the things I've been talking about. I'm sure you won't have seen the effects of changing the regulation gap in real time, so it's worth sharing what happens when you do that. I'm still to check if the pockets are necessary for the Delivery overflow, my guess is that they aren't.  Cone51 by Georgia Montgomery, on Flickr Along the way, I managed to drill the pockets for the 0.5mm holes and then forgot to drill them. Doh! Rather than waste the cone, I set it up using the drill chuck on the parallel mandrel so give me enough clearance for the spindle nose. I just put a 0.9mm drill through the holes I did drill and used a square to eye it up.  20210212_124931 by Georgia Montgomery, on Flickr This is the shape of the Steam Cone nose I've been using...  20210212_160005 by Georgia Montgomery, on Flickr ... and this is the more rounded way it is now. To be honest, I don't think it makes much difference. It's certainly not as important as the one on the cone entry.  20210212_160454 by Georgia Montgomery, on Flickr |
|
|
|
Post by delaplume on Feb 12, 2021 21:27:35 GMT
|
|
Gary L
Elder Statesman
Posts: 1,208
|
Post by Gary L on Feb 12, 2021 23:53:56 GMT
Isn't that what moving cone injectors are all about ??.........ie to automatically adjust to changing conditions ... I have no idea Alan, I imagined it was just about starting them. I guess you could contrive a setup where the gap automatically adjusted itself to be smaller as the pressure drops. It's something I have contemplated, but there isn't much room to fit anything. I wondered about a sprung sliding sleeve that's forced back when the pressure is high. The trouble with any moving parts is the issue of dissolved solids messing with the fits and clearances in our scale. I'll leave that one for someone else to experiment with. The bottom line is that an injector that works from 90-40psi without adjustment is probably plenty good enough for what we need in my opinion. If it goes a lot lower with adjustment, that's a bonus. Hi Roger If you are aiming for injectors that are optimised for your Speedy (and why not?) then that is certainly true. But if you are aiming for a more universal applicability, then bear in mind that there are established designs that run on higher boiler pressures (Bridget/Jessie 120 psi, Paddington 110 psi to mention only two that I know well. I'm sure there are plenty of others). 'Trade' injectors seldom work well at these pressures, though it is possible to get 'specials' made that will... though that invariably affects the low pressure operation adversely. Gary |
|
|
|
Post by Roger on Feb 13, 2021 9:23:05 GMT
I have no idea Alan, I imagined it was just about starting them. I guess you could contrive a setup where the gap automatically adjusted itself to be smaller as the pressure drops. It's something I have contemplated, but there isn't much room to fit anything. I wondered about a sprung sliding sleeve that's forced back when the pressure is high. The trouble with any moving parts is the issue of dissolved solids messing with the fits and clearances in our scale. I'll leave that one for someone else to experiment with. The bottom line is that an injector that works from 90-40psi without adjustment is probably plenty good enough for what we need in my opinion. If it goes a lot lower with adjustment, that's a bonus. Hi Roger If you are aiming for injectors that are optimised for your Speedy (and why not?) then that is certainly true. But if you are aiming for a more universal applicability, then bear in mind that there are established designs that run on higher boiler pressures (Bridget/Jessie 120 psi, Paddington 110 psi to mention only two that I know well. I'm sure there are plenty of others). 'Trade' injectors seldom work well at these pressures, though it is possible to get 'specials' made that will... though that invariably affects the low pressure operation adversely. Gary Hi Gary, Yes, I'm optimising this for my own locomotive. However, you only have to reduce the throat diameter of the Steam Cone to be able to use the same design for higher pressures. As it is, the design works at 100psi, and there's every chance it will work slightly higher. Both Bob Bramson and D.A.G Browm show the relationship between throat sizes for the different pressure ranges. |
|
|
|
Post by Roger on Feb 13, 2021 19:49:53 GMT
Here are the videos I promised yesterday. The first one is the testing of Cone 53 which is the best so far. Annoyingly Flickr seems to clip them at a fixed length, so I'm uploading them to YouTube so you can see the whole of the footage. YouTube link 20210213_155855 by Georgia Montgomery, on Flickr This video seeks to demonstrate the effects of adjusting the regulation gap. Hopefully it makes sense, but I'll clarify it with some more comments if it's confusing. I think you can see from this how critical it is to get the gap just right. The thread I'm adjusting is 0.75mm pitch, so a quarter of a turn is 0.1875mm (7 thou) and you can see how little I'm turning the thread to get the fine adjustment. This is more sensitive than when using Annular Regulation, so any design using this ought to have a way to quickly try out different gaps before making an appropriate shim or setting the length. It was a nightmare trying to get a camera angle and lighting that allowed reading both the gauges and getting the rest in view. You can see the bottom of the valve bobbing up and down when the overflow operates under pressure. It moves a long way but you can see that it also seals really well. Hopefully you can see enough of the overflow pipe to get a flavour of what's going on there and can hear the sounds it makes. YouTube link 20210213_162737 by Georgia Montgomery, on Flickr |
|
