|
Post by Roger on Mar 21, 2019 8:02:54 GMT
It's one of those things that is actually so simple and obvious you wonder why you didn't think of it earlier, but yes it makes perfect sense for a few reasons. I think the downside is that there's less volume for the excess steam and air to escape into. You also have that slitting operation that probably puts some people off, especially if the only have a lathe. Having said that, it's easy enough to make a fixture to hold the cones and have the slitting saw in the chuck. There's also the issue of how strong the one piece cone is. I don't think it would survive pressing into the body with anything but the lightest of interference fits. Personally, I'd make it a sliding fit and use Loctite to hold it in place. That sort of solution would probably be frowned upon by traditionalists. There's no reason to remove them when installed, so I don't see an issue with that. It's always possible to heat up the assembly and get them out if it's really necessary. Convention is probably the strongest reason that the use of a one piece cone hasn't spread. If it works as well as two separate cone, it's a compelling solution. But, as I've pointed out, there are differences in how you might successfully implement them.
|
|
|
Post by RGR 60130 on Mar 21, 2019 8:44:54 GMT
In my ancient copy of LBSC's Live Steam Book he shows the two types of combining cone. The two part he calls the Holden & Brooke type while the one part he calls the Sellers type. He suggests the Holden & Brooke is quicker to pick up while, as already agreed, the Sellers type is easier to make.
Reg
|
|
robmort
Hi-poster
3.5" Duchess, finishing 2.5" gauge A3 and building 3.5" King
Posts: 172
|
Post by robmort on Mar 21, 2019 9:39:28 GMT
This gets into the realm of philosophy as to what accuracy is needed and can be achieved for any parts we make. Even a single piece cone will have a limited accuracy e.g. surface roughness, alignment etc. What accuracy is needed for the smallest dimensions no-one knows: 0.01%, 1%? Down to judgment and experience in the end.
|
|
|
Post by Roger on Mar 21, 2019 10:13:36 GMT
This gets into the realm of philosophy as to what accuracy is needed and can be achieved for any parts we make. Even a single piece cone will have a limited accuracy e.g. surface roughness, alignment etc. What accuracy is needed for the smallest dimensions no-one knows: 0.01%, 1%? Down to judgment and experience in the end. I agree, it's not at all clear what accuracy is required. Thinking about the function of the Injector, it would be useful to know which errors have the greatest consequences. The absolute angles seem to be the least important, however the throat diameters necessary to make the design work are determined by those angles. The throat diameters appear to be critical since the ratios of them are what create the Ejector and Injector proportions. So although you could argue that it doesn't matter if you use 9 degrees or 10 degrees for example, if you design it for 9 degrees then I'd suggest that you need to get that angle as close as possible to that figure. This is far from easy. Getting everything concentric is relatively easy, you just have to machine the inside and outside of the cones in the same setup. If you have to turn a cone round and machine it from both sides, I think you need clock it up in a 4-jaw or GripTru or have a collet arrangement. Holding it in the 3-jaw isn't going to be anywhere near enough in my opinion. There's no reason why the concentricity of the cones can't be within a couple of microns with a little care. If you can't hold the body so it's running absolutely true, you can't get the hole to go in the middle. I can see how using tiny HSS drills is likely to end up with them wandering. Personally I wouldn't use them, they're too flexible. PCB carbide drills are the ideal tool for parallel holes but you need to use them carefully to avoid breakage. I doubt very much if most lathes have the drill chuck accurately centred on the axis of rotation of the chuck. I might make a special chuck with an adjustable nose so that it can be set precisely to the correct height and front to back position. I think if you put a clock in the chuck and clock around a rod held in the drill chuck, most lathes would be miles out without adjustment. You can't expect to drill deep small holes down the axis of the cones with a drill that's miles off the centre line! So, there are more questions than answers really, but I'm going to aim to make the perfect geometry, and start from that idealistic standpoint. We all know that errors are going to creep in, but being obsessed with getting it super accurate is probably the way to go. After all, you can design the geometry to be exactly how you want it, but the further you make it from that design, the less likely it is to work.
|
|
|
Post by chris vine on Mar 21, 2019 17:13:38 GMT
hi Roger,
if you make a one piece cone for the centre section and are worried about its strength when pressing it in, you could always relieve one end of it by a few microns and press it in with the other end. Then it won't tend to collapse the fragile centre spacing part.
You have thought of this already of course!!
Chris.
|
|
|
Post by Roger on Mar 21, 2019 17:38:28 GMT
hi Roger, if you make a one piece cone for the centre section and are worried about its strength when pressing it in, you could always relieve one end of it by a few microns and press it in with the other end. Then it won't tend to collapse the fragile centre spacing part. You have thought of this already of course!! Chris. Hi Chris, That's an idea, if not really considered that particular option. I think I'd be more inclined to just make it a really close sliding fit and use Loctite. It's hard to create a press fit and control the amount of force. The difference between it bring too tight and too loose is only a matter of a few microns
|
|
|
Post by ettingtonliam on Mar 21, 2019 18:08:18 GMT
Thinks: I wonder how LBSC managed to make injectors that work (and it seems to be agreed that he did) with the equipment he had, and before microns had been invented?
|
|
|
Post by 92220 on Mar 21, 2019 18:15:06 GMT
Hi Roger.
Is Loctite OK in an injector? I've no idea how hot they get. I had been toying with the idea of making the Type K injector for my loco, in bronze, and fitting the cones in a sleeve, Loctited in place, in a bore up the middle but didn't know if Loctite would be suitable. If it is Ok then it sounds a feasible proposition.
Bob.
|
|
jma1009
Elder Statesman
Posts: 5,896
|
Post by jma1009 on Mar 21, 2019 19:46:54 GMT
The only thing I can add to the combining cone gap (between the 2 halves), and I am writing this from memory rather than plough back through my file of notes is as follows:-
1. The Ted Crawford combining cone relates to a very large miniature injector (I think it was Lisa who posted the drawings a few years back on here)
2. The bore through the injector body for the cones has quite an effect on performance on all 3 gaps. John Cashmore, who commercially made injectors, would alter the bore to produce different male threads on the ends of the bodies. He may have done these as 'specials' at the request of customers to avoid them having to make new pipework and fit new cones and union nuts. I was surprised one I was sent had a reduced bore and threads on the ends of the bodies, but otherwise exactly the same as his standards. It worked fine, but when starting/picking up was somewhat 'explosive'! 'Odd' 9/32" x 40 or 32 TPI ME threads which Gordon Chiverton sometimes used, which necessitated me making up a set of new pipework
3. LBSC had an extremely well equipped workshop including a mill and his Milnes lathe, and a Boley lathe that he used for making the injector cones. He ran a precision machine shop in Weybridge during WW1. So far as his writings on injectors in ME are concerned, he deliberately watered down his own precision approach, which would pretty much result in the amateur Model Engineer in the 1940s and 1950s copying his 'Words and Music' only producing a working injector by luck! And then if it worked it would require much fiddling about with to pick up with that we would find unacceptable today. (For those who were lucky enough to be sent a personally made Injector by LBSC as a Christmas present, I understand they invariably worked quite well! Or at least no one ever went into print to complain their Christmas present did not work!)
4. For checking cone tapers, you make up a set of taper gauges - tapered 'D' bit reamers without the 'D' bit. Using say a standard 9 degree taper gauge, if it wobbles at the large end the taper of the cone is more than 9 degrees. If it wobbles at the small end it is smaller than 9 degrees. Using a set of taper gauges you can work out quite precisely the actual taper of the cones.
5. If you ream the 2 halves of the combining cone with a 9 degree taper, you can use the taper gauge with a bit of 'blue' added to see whether the 2 halves have been pressed in accurately. If you left them undersize on reaming, a 9 degree taper reamer with a body same diameter as the bore of the body, and with a stop added, can be inserted and rotated to obtain perfect concentricity, and perfect matching taper on the 2 halves.
Cheers,
Julian
|
|
|
Post by Roger on Mar 21, 2019 20:15:24 GMT
Hi Roger. Is Loctite OK in an injector? I've no idea how hot they get. I had been toying with the idea of making the Type K injector for my loco, in bronze, and fitting the cones in a sleeve, Loctited in place, in a bore up the middle but didn't know if Loctite would be suitable. If it is Ok then it sounds a feasible proposition. Bob. Hi Bob, I can't see temperature being a problem, the body is being cooled by water unless there isn't any being fed. If that's the case, the absolute maximum it could be would be 166C ie the boiling point at 90psi. Since there's going to be a loss in pressure before it reaches the injector, plus the cooling effect of it expanding in the steam cone, it's unlikely it will be anything like that temperature. A typical Loctite, say 601, has a maximum working temperature of 150C which is almost certainly sufficient. If not, high temperature ones go much higher.
|
|
|
Post by Roger on Mar 21, 2019 20:28:31 GMT
The only thing I can add to the combining cone gap (between the 2 halves), and I am writing this from memory rather than plough back through my file of notes is as follows:- 1. The Ted Crawford combining cone relates to a very large miniature injector (I think it was Lisa who posted the drawings a few years back on here) 2. The bore through the injector body for the cones has quite an effect on performance on all 3 gaps. John Cashmore, who commercially made injectors, would alter the bore to produce different male threads on the ends of the bodies. He may have done these as 'specials' at the request of customers to avoid them having to make new pipework and fit new cones and union nuts. I was surprised one I was sent had a reduced bore and threads on the ends of the bodies, but otherwise exactly the same as his standards. It worked fine, but when starting/picking up was somewhat 'explosive'! 'Odd' 9/32" x 40 or 32 TPI ME threads which Gordon Chiverton sometimes used, which necessitated me making up a set of new pipework 3. LBSC had an extremely well equipped workshop including a mill and his Milnes lathe, and a Boley lathe that he used for making the injector cones. He ran a precision machine shop in Weybridge during WW1. So far as his writings on injectors in ME are concerned, he deliberately watered down his own precision approach, which would pretty much result in the amateur Model Engineer in the 1940s and 1950s copying his 'Words and Music' only producing a working injector by luck! And then if it worked it would require much fiddling about with to pick up with that we would find unacceptable today. (For those who were lucky enough to be sent a personally made Injector by LBSC as a Christmas present, I understand they invariably worked quite well! Or at least no one ever went into print to complain their Christmas present did not work!) 4. For checking cone tapers, you make up a set of taper gauges - tapered 'D' bit reamers without the 'D' bit. Using say a standard 9 degree taper gauge, if it wobbles at the large end the taper of the cone is more than 9 degrees. If it wobbles at the small end it is smaller than 9 degrees. Using a set of taper gauges you can work out quite precisely the actual taper of the cones. 5. If you ream the 2 halves of the combining cone with a 9 degree taper, you can use the taper gauge with a bit of 'blue' added to see whether the 2 halves have been pressed in accurately. If you left them undersize on reaming, a 9 degree taper reamer with a body same diameter as the bore of the body, and with a stop added, can be inserted and rotated to obtain perfect concentricity, and perfect matching taper on the 2 halves. Cheers, Julian Hi Julian, Thanks for clarifying what size that one piece combining cone came from. Although the slot is smaller in a miniature injector, it's certainly feasible to cut that with a slitting saw. I have one that's 0.35mm wide which ought to be about right for the sizes I'm considering. Cutting away the clearance material would take a little thought, I wouldn't want to be getting in there with needle files. I think that could be machined with a dovetail cutter. I think making precision taper gauges is beyond most Model Engineers. Getting the angle precisely right is not a trivial matter, and checking what you've made is even more difficult. All of these things can be done, but how many bother? It's the same problem with making the taper reamers in the first plate. Again, knowing the precise position of a stop on the reamer is going to take some care. I certainly agree that there are ways to get the alignment to be very close when making the cones separately. It does beg the question as to why you would do that when it's guaranteed to be aligned perfectly if you make them in one piece? At the moment I'm undecided on this, but it's certainly tempting to go down the one piece route.
|
|
jma1009
Elder Statesman
Posts: 5,896
|
Post by jma1009 on Mar 21, 2019 21:22:49 GMT
Hi Roger,
You have the kit to experiment and use different methods.
Making a correct taper gauge is just a bit of maths and trig and machining to suit.
I had 2 people who influenced me when I was a teenager; Arthur Grimmett, and Ron Shephard founder member of the IWMES. Ron burnished his cones using these taper gauges out of silver steel hardened and tempered and dead smooth. Arthur made pentagonal high speed steel tapered reamers on his grinder that John the Pump now has (both reamers and grinder). Arthur made a number of tapered gauges as well, though these may just have been blanks, as I don't think Arthur ever used them.
But I am veering off into how to make them as opposed to designing them!
The explosive characteristics of the smaller bore John Cashmore injectors when picking up was the point I wanted to emphasise!
Cheers,
Julian
|
|
jma1009
Elder Statesman
Posts: 5,896
|
Post by jma1009 on Mar 21, 2019 23:40:53 GMT
As an aside, as to how to make them as to how to design them, Arthur Grimmett always re-reamed his combining cones when inserted into the injector body. Don Young's 'County Carlow' injector description of manufacture in ME 1970 is a copy of how Arthur did it, and I have not the slightest doubt that Gordon Chiverton did the same for ever after till Gordon died.
Adam Cro and Paul will know whether or not Len Steel does the same as I have assumed for Gordon Chiverton.
If I remember correctly in that halcyon summer school holiday of 1983, when Don Young had a big 'fall out' with Gordon Chiverton, and got Arthur out of retirement to make batches of injectors to fulfill Don's orders stacking up, Arthur had only to re-ream the combining cones in 2 parts "a whisker" which I think is the same phrase LBSC used. I was in Arthur's famous 'Aptule Works' at Apse Heath Isle of Wight that summer holiday in the weekdays.
Arthur picked out one of his best working mediuum standard sized injectors to fit to one of his locos, and when I examined it many years later I was rather surprised it had a very generous combining cone throat similar to the Gordon Chiverton proportions!
Cheers,
Julian
|
|
|
Post by Roger on Mar 22, 2019 7:06:24 GMT
Hi Julian, It certainly makes sense to fine ream the cones when they're in the body, but the fact that this is done does indicate that the alignment is probably not as good as it would ideally be.
If you have a grinder then that's the obvious tool to use for making the reamers. These days it's even possible to make them from Carbide if you so wish. Personally, I think HSS is fine for this application and is less likely to break. I'm fortunate enough to be able to set any angle to very high precision, but although I agree that it's only a bit of trigonometry to set the angle, doing it accurately takes a bit of care. There are some good YouTube videos on this though.
I can see that a gauge might be useful if you're trying to find out whether the two combining cones are aligned, but there's little point if they're made in one.
When I get a bit further, I might start a new thread exclusively about making injectors since it's an interesting subject worthy of a lot of discussion.
|
|
robmort
Hi-poster
3.5" Duchess, finishing 2.5" gauge A3 and building 3.5" King
Posts: 172
|
Post by robmort on Mar 22, 2019 10:40:59 GMT
........ you could argue that it doesn't matter if you use 9 degrees or 10 degrees for example, if you design it for 9 degrees then I'd suggest that you need to get that angle as close as possible to that figure. This is far from easy. I've seen no evidence that the cone angles are critical to within a degree. Different designs use different values and with unknown accuracy. What matters most are the relative orifice dimensions which determine the flow rate and performance. I don't think it's a problem with flexible drills or off-axis drill chucks. If you drill a long hole even with a much larger drill, say 4mm, over a 10x depth (40mm) with minimal off-axis you often find that the drill has wandered off centre at the end unless you're very careful, because of swarf, uneven flutes or just because that what it likes to do. Slow feed and taking little "bites" helps. But it's most important to start the drill accurately on-centre with a special shaped centre drill.
|
|
|
Post by Roger on Mar 22, 2019 11:25:00 GMT
........ you could argue that it doesn't matter if you use 9 degrees or 10 degrees for example, if you design it for 9 degrees then I'd suggest that you need to get that angle as close as possible to that figure. This is far from easy. I've seen no evidence that the cone angles are critical to within a degree. Different designs use different values and with unknown accuracy. What matters most are the relative orifice dimensions which determine the flow rate and performance. I don't think it's a problem with flexible drills or off-axis drill chucks. If you drill a long hole even with a much larger drill, say 4mm, over a 10x depth (40mm) with minimal off-axis you often find that the drill has wandered off centre at the end unless you're very careful, because of swarf, uneven flutes or just because that what it likes to do. Slow feed and taking little "bites" helps. But it's most important to start the drill accurately on-centre with a special shaped centre drill. Hi Rob, I'm not saying that the angles are critical. If you read what I said again, you'll see that I'm suggesting that if the design calls for 9 degrees, you want to make sure you use a 9 degree reamer. If the angle is different to the design, you won't get the right throat diameters in the two parts of the combining cone. I agree that good centring is essential. The longer the hole, the more chance of it wandering but I'd wager that the more off centre the tailstock is, the more it will wander, even if the centring is spot on. Quite how you would centre accurately with a tailstock that's off centre is unsure. The point I'm making about PCB drills is that they are ground to very high tolerances indeed. If you look at a 0.1mm PCB drill under a microscope, it looks almost perfect. I very much doubt is regular HSS drills being like that. It's necessary for them to be that accurate because they work in a hellish environment at speeds unimaginable in general engineering. A 0.1mm drill will typically be used at 250,000RPM Larger PCB drills are ground on the same equipment to the same tolerances, so using those to start the hole for an injector makes sense. Regularly withdrawing the drill definitely makes sense.
|
|
|
Post by andyhigham on Mar 22, 2019 11:44:57 GMT
When drilling deep holes, if possible I drill pilot holes from either end. If the drill wanders the two holes will not line up but the next drill through will get pulled onto the correct trajectory (within reason)
|
|
robmort
Hi-poster
3.5" Duchess, finishing 2.5" gauge A3 and building 3.5" King
Posts: 172
|
Post by robmort on Mar 22, 2019 14:23:10 GMT
.....I'm not saying that the angles are critical. If you read what I said again, you'll see that I'm suggesting that if the design calls for 9 degrees, you want to make sure you use a 9 degree reamer. If the angle is different to the design, you won't get the right throat diameters in the two parts of the combining cone. Yes I understand and agree with what you said, but there's another way to adjust the position of the cones to match the actual cone taper, which I've done. That is to measure the cone wider-end diameter (e.g. with a drill shank) and then use suitable spacers between the cones to set the gaps. This allows experimentation with cone gaps without machining new cones each time. Even if the tailstock is centred, the drill may be bent slightly or simply off-centre. But a spotting or centre drill is usually self-centring so doesn't need to be perfectly on-axis, and I often use a self-centring technique that works with any ordinary drill without first centring. I also have a 1000 grit diamond plate to sharpen drills under magnification.
|
|
|
Post by Roger on Mar 22, 2019 15:49:11 GMT
.....I'm not saying that the angles are critical. If you read what I said again, you'll see that I'm suggesting that if the design calls for 9 degrees, you want to make sure you use a 9 degree reamer. If the angle is different to the design, you won't get the right throat diameters in the two parts of the combining cone. Yes I understand and agree with what you said, but there's another way to adjust the position of the cones to match the actual cone taper, which I've done. That is to measure the cone wider-end diameter (e.g. with a drill shank) and then use suitable spacers between the cones to set the gaps. This allows experimentation with cone gaps without machining new cones each time. Even if the tailstock is centred, the drill may be bent slightly or simply off-centre. But a spotting or centre drill is usually self-centring so doesn't need to be perfectly on-axis, and I often use a self-centring technique that works with any ordinary drill without first centring. I also have a 1000 grit diamond plate to sharpen drills under magnification. Surely the design is for a particular angle, and this is what defines the ratio of the different throat diameters. You can move the Steam cone to adjust the regulation, but if the angle of the combining cones are different to the design value, either the Condensing cone or Mixing cone throat sizes will be compromised. Ok, you might be able to get it to work by moving things around, but I think it's more sensible to make it as close to the drawing as possible rather than try to fix something that's wrong. I think it's this tinkering to get things working that confuses the whole issue of what's right and what's wrong. However close you get the centre, surely it makes sense to take the trouble to align the drill chuck with the centre line? How big an error do you think is acceptable? I don't think you would get even close to what's achieved on a PCB drill sharpening machine, even with a tool and cutter grinder, let alone by hand! We're talking about exquisitely ground micro-grain carbide here, it's in a different league to most commercial general purpose drills. They cost around £1 a piece, why would you want to sharpen something by hand and risk scrapping the job?
|
|
|
Post by Oily Rag on Mar 22, 2019 20:21:15 GMT
Clipped for clarity, Roger wrote "I don't think you would get even close to what's achieved on a PCB drill sharpening machine, even with a tool and cutter grinder, let alone by hand! We're talking about exquisitely ground micro-grain carbide here, it's in a different league to most commercial general purpose drills. They cost around £1 a piece, why would you want to sharpen something by hand and risk scrapping the job?"
It is because most of us know nothing of PCB carbide drills etc and their great advantages. I was ignorant till I learnt from your posts on these forum pages and I then tried them out after buying some via ebay. Yep, I agree with what you say. Brilliant tools. I cannot recall them mentioned in the pages of ME mags etc. So I think it is all that we know after decades of reading these publications and or the handing down of techniques from the senior makers thus the recognition of new approaches/techniques takes time to be outed and then absorbed. If LBSC had carbide PCB drills he would have surely spoken in praise of the them.
|
|