|
Post by Roger on Aug 31, 2020 7:56:23 GMT
The injector bodies are going to be machined in two halve which will be Silver Soldered together, the same way I did the 90 degree elbow. The lugs are 0.7mm thick and have holes for M1.4 bolts and 0.8 rivets. These will be machined from 2" diameter PB104 bar, but it's a really tight squeeze to fit in the diameter. I'm going to get the bar first, and maybe tweak the design if it won't quite fit. The next size up for bar is 2-1/2" and that's almost twice the price. I've ordered a 150mm length for £45.49 which is cheaper than buying one commercial injector, so it's not that expensive really. I'm using Phosphor Bronze because the curved channel is really thin so as to be able to get a large bore diameter. The only other option is Leaded Bronze SAE660, but there might be porosity issues with that after Silver Soldering. I've also got from pretty small threads to go in the flanges which will be much stronger with PB104. The flanges ought to have clearance holes, but getting fingers to the really awkward places to get nuts on them is going to be a pain. Instead, I'll tap them, and then fit cosmetic nuts on the ones that can be seen and more easily reached. So the plan is to face off the bar and machine the shape shown below in the end of it. I've created two configurations of the Injector model in the one 3D modelling file. One is the complete body to finished sizes, the other also cuts it in half and adds the lugs. I've also changed the bore diameters so there's a little to be finish machined with the exception of the curved channel which obviously can't be reached. Inside half by Roger Froud, on Flickr I then plan to part it off which will completely release the shape from the 2" diameter. I'll then counterbore the 1.4mm hole positions to take cap screws and bond the part to a fixfure with Superglue to assist the bolts. I'll use 2mm cutters to 3D machine the outside, using a tool path with keepouts where the lugs are. Finishing will be done with a 1mm ball nosed cutter. When both halves are finished, they will be riveted together and Silver Soldered. Then the lugs can be cut off and hand finished. The bores will need a fixture to hold the assembly for machining. The mirrored halves will use my utility program to flip the X-axis coordinates, I won't need to generate a different tool path with the CAM package. Outside half by Roger Froud, on Flickr Anyway, I can't see any show stopping issues with this, but that doesn't meant that there aren't any! What can possibly go wrong... At least I've got 150mm of bar to play with, so I can have more than one go.
|
|
|
Post by Roger on Aug 31, 2020 18:25:36 GMT
I've been chatting with Adam today about 10X injectors and he spotted that mine look different to his. I just assumed that the 10X had gone through different iterations and that mine was just a different casting. However, looking more closely at my 1501 pictures and zooming in on one injector, it's now possible to read 8X on the side.
So, the question is, does anyone have the Works Drawing for the 8X injector? If not, does anyone know what the difference is?
|
|
|
Post by delaplume on Sept 1, 2020 8:54:50 GMT
Hi Roger,
If I remember correctly from my time at SVR machining the full size ones-------they share the same body casting, it's just a different cone /cones that give the greater output for a 10X..........Sometimes you might see for example the 8X stamping cancelled out and 10X used ....
Please cross reference this with another source as it's over 30 years now and memory can play tricks..... Hopefully I'm correct and you only have to produce the one unit.....
PS}--- the more observant may have noticed that my 1366 class loco is missing it's injectors entirely......It should be the same set-up as 1501 ie.. under the rear of the main water tanks......Does anyone provide a solid dummy unit just for show ??
Thanks Roger
Alan
|
|
|
Post by Roger on Sept 1, 2020 10:01:22 GMT
Hi Roger, If I remember correctly from my time at SVR machining the full size ones-------they share the same body casting, it's just a different cone /cones that give the greater output for a 10X..........Sometimes you might see for example the 8X stamping cancelled out and 10X used .... Please cross reference this with another source as it's over 30 years now and memory can play tricks..... Hopefully I'm correct and you only have to produce the one unit..... PS}--- the more observant may have noticed that my 1366 class loco is missing it's injectors entirely......It should be the same set-up as 1501 ie.. under the rear of the main water tanks......Does anyone provide a solid dummy unit just for show ?? Thanks Roger Alan Hi Alan, That can't be right, because the castings look different. That's what threw up the query in the first place. The below picture from the Bluebell Railway web site is labelled as a 10X type, and it does match the 10X Works Drawing in the way the body casting necks down next to the outlet clack. 10X injector by Roger Froud, on Flickr Contrast that with this picture of an 8X injector fitted to 1501, where the large diameter of the body continues past the outlet clack to the end. There are other visual differences, such as the diameter of the body at the inlet end and the additional web under the water inlet tube which extends to the body. DSCN5687 by Roger Froud, on Flickr Maybe someone knows of the history of these types. Are they part of a 1X-10X series of designs? That seems unlikely since 15xx was designed right at the end of the Steam era. However, is it possible that 8X and 10X types are interchangeable and that 1501 may have been originally fitted with 10X types? Maybe the number relates to the capacity? Who knows. As far as I can tell, the 8X and 10X are approximately the same size, judging by the fact that the flanges look to be in similar proportions to the body. So what I've done is to model the injector on the 10X Works Drawing, but changed the ends to reflect how it looks on 1501. That's going to be good enough, but I'm curious to know a bit more of their history and why they're different even though they appear substantially the same.
|
|
|
Post by Roger on Sept 1, 2020 10:01:40 GMT
Hi Roger, If I remember correctly from my time at SVR machining the full size ones-------they share the same body casting, it's just a different cone /cones that give the greater output for a 10X..........Sometimes you might see for example the 8X stamping cancelled out and 10X used .... Please cross reference this with another source as it's over 30 years now and memory can play tricks..... Hopefully I'm correct and you only have to produce the one unit..... PS}--- the more observant may have noticed that my 1366 class loco is missing it's injectors entirely......It should be the same set-up as 1501 ie.. under the rear of the main water tanks......Does anyone provide a solid dummy unit just for show ?? Thanks Roger Alan Hi Alan, That can't be right, because the castings look different. That's what threw up the query in the first place. The picture below, from the Bluebell Railway web site, is labelled as a 10X type, and it does match the 10X Works Drawing in the way the body casting necks down adjacent to the outlet clack. 10X injector by Roger Froud, on Flickr Contrast that with this picture of an 8X injector fitted to 1501, where the large diameter of the body continues past the outlet clack to the end. There are other visual differences, such as the diameter of the body at the inlet end and the additional web under the water inlet tube which extends to the body. DSCN5687 by Roger Froud, on Flickr Maybe someone knows of the history of these types? Are they part of a 1X-10X series of designs? That seems unlikely since 15xx was designed right at the end of the Steam era, and it has an 8X type. However, is it possible that 8X and 10X types are interchangeable and that 1501 may have been originally fitted with 10X types? Maybe the number relates to the capacity? Who knows. As far as I can tell, the 8X and 10X are approximately the same size, judging by the fact that the flanges look to be in similar proportions to the body. So what I've done is to model the injector on the 10X Works Drawing, but changed the ends to reflect how it looks on 1501. That's going to be good enough, but I'm curious to know a bit more of their history and why they're different even though they appear substantially the same.
|
|
|
Post by 92220 on Sept 1, 2020 18:02:00 GMT
Hi Roger.
Those photos are intriguing. Evening Star is fitted with a 10X Restarting Injector which exactly matches the works drawing SL/SW/415, I have, but neither of your photos has the same casting. I will try and get a photo uploaded. I am guessing your castings are just upgrades, to the base design, over time.
Bob. Edit: I have just looked at the Swindon drawing of the Overflow Valve. It is standard on 8X, 10X and 11X injectors, according to the drawing, so I would guess that the number just refers to the cones and output of each type of injector, as you thought, and the main casting is standard for all of them.
|
|
|
Post by dhamblin on Sept 1, 2020 18:58:58 GMT
The injector surfaces look very clean with few scratches or dings, which makes me think it is a relatively new preservation era casting. In fact, looking at the right hand water valve you can see it is a new casting with 'SVR' on it, which is most likely Severn Valley Railway. So on balance it is probably cast from a new pattern made by the railway.
Regards,
Dan
|
|
|
Post by delaplume on Sept 1, 2020 21:59:37 GMT
Hello all, Bob, as per your "edit"....this is what I had in mind earlier on when I said that the body is standard and the output varied by having different cones etc...no doubt the body externally may have been "Tweaked" by Swindon in the course of time and thus operational experiences... Hello Dan......yes, I can confirm that the legend SVR on the water valve casting does indeed refer to the Severn Valley Railway...I myself having overhauled old and machined new ones during my time there......In those days the Pattern Making facility was in it's early days with my then Brother - in Law Brian Oldford being one of the Pattern Makers....Other items produced included Hydrostatic Lubricator reservoir body and loco clack valve bodies............... www.gw-svr-a.org.uk/bridgnorth_pattern_shop_2007.html ................. If you go to the injector section in The Black Book fig. 24 "Injector combining cone with hinged flap" and look at the R/H side ( lower half ) and you will see a web indicated as in the Bluebell Railway photo.....The diagram was produced by British Railways ( Western Region )... Hello Roger, If all of your proposed dimentions will fit inside the body as worn by 1501 then why not do it like that ??.....It will also follow as part of the " Build it like it is at the moment" ethos as well....
|
|
|
Post by Roger on Sept 2, 2020 6:34:51 GMT
Hello all, Bob, as per your "edit"....this is what I had in mind earlier on when I said that the body is standard and the output varied by having different cones etc...no doubt the body externally may have been "Tweaked" by Swindon in the course of time and thus operational experiences... Hello Dan......yes, I can confirm that the legend SVR on the water valve casting does indeed refer to the Severn Valley Railway...I myself having overhauled old and machined new ones during my time there......In those days the Pattern Making facility was in it's early days with my then Brother - in Law Brian Oldford being one of the Pattern Makers....Other items produced included Hydrostatic Lubricator reservoir body and loco clack valve bodies............... www.gw-svr-a.org.uk/bridgnorth_pattern_shop_2007.html ................. If you go to the injector section in The Black Book fig. 24 "Injector combining cone with hinged flap" and look at the R/H side ( lower half ) and you will see a web indicated as in the Bluebell Railway photo.....The diagram was produced by British Railways ( Western Region )... Hello Roger, If all of your proposed dimentions will fit inside the body as worn by 1501 then why not do it like that ??.....It will also follow as part of the " Build it like it is at the moment" ethos as well.... Hi Alan, Yes, that's what I plan to do. I was just a bit concerned that the 8X body might have been smaller externally.
|
|
baldric
E-xcellent poster
Posts: 210
|
Post by baldric on Sept 2, 2020 6:55:23 GMT
The 8x & 10x injectors I think are very slightly different in length, how do I know this? Well many engines had 2 different sizes, on 7202 it has one of each, as do the 41/51xx, and the pipework is slightly different. I do not currently have a picture to show you the difference. I believe the 8 & 10 refer to the nozzle diameter in millimetres. There is a drawing of one of the injectors on display at Didcot on the loco-shed wall, I can't remember which one it is though. There are smaller sizes, but it may only go down to 6x. There is a good image of an 8mm injector here
|
|
|
Post by Roger on Sept 2, 2020 7:48:07 GMT
The 8x & 10x injectors I think are very slightly different in length, how do I know this? Well many engines had 2 different sizes, on 7202 it has one of each, as do the 41/51xx, and the pipework is slightly different. I do not currently have a picture to show you the difference. I believe the 8 & 10 refer to the nozzle diameter in millimetres. There is a drawing of one of the injectors on display at Didcot on the loco-shed wall, I can't remember which one it is though. There are smaller sizes, but it may only go down to 6x. There is a good image of an 8mm injector here Hi Baldric, Thanks for that. I'll probably slightly shorten the body then by about 0.2mm to be able to get it more comfortably out of the 2" Phosphor Bronze bar. It won't show, and it might actually be closer to what's on 1501. The main thing is for it to look right, small discrepancies I can live with.
|
|
|
Post by 92220 on Sept 2, 2020 7:48:44 GMT
|
|
|
Post by Roger on Sept 2, 2020 15:26:05 GMT
Hi Bob, Yes, those definitely look like 10X injectors. The ones on 1501 now are definitely 8X, so that explains the difference in the castings.
|
|
JonL
Elder Statesman
WWSME (Wiltshire)
Posts: 2,975
|
Post by JonL on Sept 3, 2020 13:35:47 GMT
I hope you are getting the brasso out Roger...
|
|
|
Post by Roger on Sept 3, 2020 20:49:34 GMT
I've not been idle on the painting front, but it's been slow progress. I wasn't happy with the awful finish on these brackets, so I carefully stripped them back to the bare metal to have another go. They're still not amazing, but I can live with it. 20200903_174845 by Roger Froud, on Flickr 20200903_174851 by Roger Froud, on Flickr However, while that's going off, look what arrived! I'm rather pleased with this, it's nice and chunky and pretty well made. 20200902_171558 by Roger Froud, on Flickr Anyway, I don't want all the gubbins it came with, I just want the basic spindle and body arrangement... 20200902_172325 by Roger Froud, on Flickr ... like this. 20200902_172928 by Roger Froud, on Flickr So I've modelled enough to be able to sort out the motor mount which ends up like this. The idea is to be able to fit it on either side because sometimes parts of the quill or high speed spindle get in the way in one orientation. Motor mount assembly by Roger Froud, on Flickr Before going too far, I thought I'd check out the Servo interface and fit the 4-th axis Servo Amplifier which has been sitting under the bench for about 6 years, still in its original packaging. I've now fitted the power cables and just need to connect the Step & Direction signals, the Enable and the 24V brake connections. 20200903_094624 by Roger Froud, on Flickr I've also taken a look at Mach4 to see what's required of that. The output signals for Step & Direction are the defaults indicated in the breakout smoothstepper board CM106ESS manual so that's easy. What wasn't quite as obvious is the number of steps required for 1 degree. The A-axis reads as a decimal number, the intention being that 1.0 is 1 degree, to as many places of decimals possible for the Servo resolution. Originally I'd used the same 30 tooth pulley arrangement that I've got on the Z-axis, and used a 90 tooth pulley on the chuck end to give 3:1. However, when you do the maths, it doesn't work out neatly. The SureServo outputs 10,000 steps per turn, so that would be 30,000 steps for 360 degrees therefore 83.33333 steps per degree. I see that other people have just typed that figure into Mach4 and it seems happy, but I don't like it. So I had a play with some other ratios, and found that 3.6:1 ends up with too large a pulley at the chuck end with the 30 tooth at the small end. However, with a bit of careful design, I've managed to shoehorn a revised taper locking arrangement on the motor end (it has a 19mm smooth shaft) into a 25 tooth pulley. That combined with a 90 tooth pulley gives the magic 3.6:1 resulting in exactly 100 steps/degree. That's more like it! So the resolution is 1/100th of a degree, or 36 seconds of arc, so that's not too bad.
|
|
dscott
Elder Statesman
Posts: 2,440
|
Post by dscott on Sept 3, 2020 23:49:07 GMT
Impressive Roger. I have a rusty one hidden somewhere without a chuck given by a Student years ago! I must make a Warco Fitting for it. The Myford can screw onto the no 2 Morse in the middle. Via a drawbar and their fitting.
I have got all my belts and pulleys from Bearing Boys. They even do Hiwins and Rails cut to length. they come in 4 M lengths so any combination out of that. Saves odd bits hanging about and is cheaper.
David and Lily.
|
|
|
Post by jon38r80 on Sept 4, 2020 12:02:30 GMT
Seems to me that you could quitre easily drive the tilting mechanism while you are at it as Wilf (UUU) suggested, or are you at the limit of your controller capacity
|
|
|
Post by Roger on Sept 4, 2020 12:52:31 GMT
Seems to me that you could quitre easily drive the tilting mechanism while you are at it as Wilf (UUU) suggested, or are you at the limit of your controller capacity I don't think it's something I would ever use to be honest. The controller will handle it, but when you consider that each new AC Servo axis costs over £1000 then I can't justify it. I bought the one for the 4th axis long ago.
|
|
|
Post by Roger on Sept 4, 2020 12:56:22 GMT
Here's the geometry for a single gap between the teeth of the 5mm pitch belt, with the three tool paths superimposed on it. I could go to the trouble of making a form tool and using it like a gear tooth cutter, but since there's only one of these, I thought I'd just 3D machine each tooth since it's easy enough to do with a single 1mm ball nosed cutter. It won't be quick, but it's not important. One tooth by Roger Froud, on Flickr
|
|
|
Post by miketaylor on Sept 4, 2020 13:57:53 GMT
Roger,
How does that "flat" tooth diagram you have produced adapt itself to the curvature of the sprocket??
Mike
|
|