I made some steam operated sanders for Bongo and discovered something important!
They worked perfectly on a test boiler on afternoon but the next day, when a visitor appeared in the workshop, they wouldn't work at all. The sand had got damp in the trap.
The reason was that the steam valve didn't shut off perfectly and so a tiny wisp of steam was still being delivered. This drifted up the sand delivery pipe and made the sand in the trap damp. The cure was to make a valve with a ptfe seat. (An o ring seat would be fine too.)
As well as playing around with sand today I have also made a start on the trap shields, since these look like castings full-size they look like a good subject for those among you who have a 3D printer.... Being a pauper myself I had to make do with some brass shim, first job was to make a pattern to shape the shim over. The picture shows the result of a few hours playing around with the lathe and mill taking measurements from the traps to judge sizes. The centre section is from copper and was made a drift fit into the brass round bar which as you can see is partly machined with a 30-degree angle to match the trap. It was then put on the rotary table to have some removed off both front and rear faces leaving the area where the tabs are, a little hand filing gave us what we see here.
I then cut up some brass shim and drilled two 1.8MM holes for the mounting tabs and allowing for the raised section that needed to be formed, this section btw is to allow air ingress into the trap as the vacuum draws the sand/air through. I guess if no air can get in it's possible for the vacuum to just crush the sandbox. I recall when making the vacuum reservoir that Don stated to test it at 100psi, of which I'm happy to say it passed, that seems an eternity ago now.
After drilling the holes, the shim was heated cherry red and formed around a round bar and then bent out using flat-faced pliers. It was then bolted to the pattern and held in an old machine vice for the next round of heating.
This is after a few more heating sessions, I did perhaps another two or three after this before calling it a day and then set about with a cutting disk and files. I used a small hammer and blunted chisel for doing most of the forming
And this is where I got to for tonight, there's a couple of blemishes which I can either live with or add a little soft solder to fill them in, I'll see how long it takes me to get the other shield to this stage before deciding on that.
assuming I get the other shield done in good time tomorrow morning I'll then make a start on the sand pipes, I'm looking forward to those...
With the shields finished I can now move on to the 'ejectors' first a couple of pictures to show where we are with the sandtraps...I have given the traps a coat of Eastwood's radiator enamel, the same paint as used on the middle cylinder which requires no undercoat and has a very fine spray. Also in the picture are the two flanges ready for the 1/8th copper sandpipe.
Here are the traps bolted to their respective sandboxes.
Next, I silver soldered the flanges to two overlength sections of 1/8th copper pipe. Prior to doing this, I did a quick test in filling a length of pipe with sand, sealing either end and bending it to an 'S' shape. This went without incident suffering from zero kinks, however, I decided to anneal the pipe anyway as I plan to bolt each sandbox to the frames and then after bolting the pipe flange to the trap bend each pipe to shape. Annealing the pipe first will allow me to do this much easier and thus impart little strain on the traps. The picture shows the flanges silver soldered in place and the pipe annealed. I have deliberately placed each flange a little way down the pipe and will then file the pipe flat so that I get a good seal when bolting in place.
I now moved on to the ejectors, Don states that his ejectors not only work but work very well, I, therefore, do not need to change the basic principle, but changes I have made in both manufacturer and look in my aim to make them look as close to the prototype as practical and be easier to maintain. To begin, here's Don's drawing...
And here's a picture of the full-size main driver sanding ejector that I took in 2016, the steam sanding gear used today is different to that of the '30s but the principle and general layout are the same.
I will begin with the ejector nozzle and as can be seen in this picture I intend to have the ejector removable as with the full-size rather than braze it all up as Don has drawn, this is both for being able to service it if required and also to make it look more like the prototype. The picture shows that I have turned down some suitable brass hex to accept a 3/16 x 40 union and ferrule. I played around with the depth of the spigot and the depth of the cone depth until I was happy with the look. The part was then drilled to accept the 1/16 pipe that will be used for the steam supply and also to make the nozzle jet.
I took this picture to give some idea of the ejector internals, the 1/16 pipe has been pulled up higher than it will be when finished just so that I can show what's going on inside. BTW, this short section of 1/16 pipe will be swaged and held in place by the cone fitted the steam pipe of the same size, hope that makes sense?
The last picture for tonight shows the two ejector nozzles and the sand pipes attached to their flanges and having been polished ready for silver soldering on the ferrule when I have decided on the pipe length and of course not forgetting to include the union nut. I did make one cockup, I had found that the length of the nozzle and the hex section that I had included to match the full size sat nicely in the chuck with the hex being at the right position to sit in one of the jaw recesses. Alas when holding the first nozzle in reverse so that I could drill the recess for the 5/32 copper pipe which fits into the end, I didn't tighten the chuck enough which duly spun and removed my little section of hex...lol I didn't make this mistake with the second nozzle, I may just put this down to one of those things as the hex isn't that prominent anyway or I may look at added it later, we shall see.
The next part to do is the sand feed, again I will have this with the union/nut as seen on the prototype which will be slightly larger at 1/4 x 40 and needs to fit the nozzle at the 35-degree angle that Don has drawn.
Hope to show that in the next day or two, thanks for looking in...
Continuing with the ejectors, I now needed to make the sand feed that will need to be silver soldered to the nozzle. As with the nozzle I have followed the prototype and included a nut and union for joining the sand pipe to the ejector. This was basically the same method as with the nozzle but now with a larger diameter body, in fact, I have made the body from much wider material than required, I'll cover the reason why shortly. The picture shows that I have turned down to a 1/4" spigot which was then threaded 1/4 x 40 TPI, the length of which was determined by testing how close I could get the nut to the body and adjusting accordingly.
After parting off, I moved onto the tilting vice and machined an angle, Don states 35-degrees, (55 from this position) but I needed to allow for clearance for the union nuts and thus settled for 45-degrees. In hindsight I should have drawn up what I wanted here as a little more length for the nozzle would have been more beneficial and closer to the prototype, I followed Don's overall dimensions which means I was limited in choosing the angle. Live and learn as they say.. The flat edge on the front is to butt up against the bracket flange, both bracket and flange being extras to what Don has drawn.
This gives us this pile of bits ready for assembly. Note that the bracket flanges have also joined the gang, I have drilled these offset as the photo looks like there is more to the flange on the top face than bottom, it could just be a trick of the camera but this will work, right or wrong.
It was then time to silver solder the nozzle to the sand feed body. I think the picture shows my method for doing this, you can also see why I machined a small part off the front of the sand body to give it a better location against the bracket flange. I'm sure that in real life this flange will be separate but I have silver soldered them to the ejector as it should help hold the ejector in its correct position, ie.. less chance of the ejector moving in service. As can be seen, I have used a short length of copper tube to hold the two body parts inline for heating, luckily this worked well. BTW, in the real size photo, it can be seen that the flange and bracket are joined by two bolts and nuts, for ease I have threaded the flange 10BA so that I don't need to have the nut on it's back which would be a little tight with the angle used between the two body's.
With those parts now joined I began to form the shape and before I got too far I remembered that I hadn't yet drilled through the sand feed body into the nozzle. For this, I played safe and inserted two drills that were a good sliding fit into the nozzle from both ends and drilled through the sand feed body until I could see the top drill move. I then took the ejector out of the lathe, removed both drills and continued drilling the hole by hand to avoid damaging the inside or worse still going right through.
It took me some time to shape the ejectors to how I wanted and having now looked at this closeup picture I can see that I still have a little left to do but they aren't that far off. As can be seen, I have blended the parts into each other to make it look like a single casting, this was the reason for making the sand feed much larger than required so that I could grind, file and sand it to blend into the nozzle. I will leave the ejector unpainted except for the nozzle outlet as per prototype.
Next job was to shape/cut the sand pipe to size and to silver solder the nozzle on. For this, the pipes were filled with sand and sealed either end. I tested this first with one sandbox bolted to the frames as seen in this next picture. Two things to note, I later reshaped the curve more and also that here the ejector is in the wrong position, what I mean is that the steam feed should be inboard, not the sand feed. This was just how it went naturally with the angle of the parts, I think though that I can manipulate this to be in the correct position, the parts are only loosely bolted together here and I'll need to have them fully secured before trying.
Lastly, a picture to show both sandboxes at the same stage, for the second I just mirrored the first, no need to turn the chassis around to fit the other sandbox if I can avoid it.
Tomorrow I'll finish off shaping the ejectors, fit the sandboxes securely to the chassis and look at how best to shape the pipes. I'll need to be careful here as they are no longer filled with sand. Once I am happy with the ejector position I'll take a look at fabricating the brackets. Not sure what they bolt too but looks like they go to the sandbox itself which if so makes life much easier. My plan is to bolt them to the nearside flange bolt that holds the sand pipe to the trap, I'll hopefully, show what I mean in the next update but I think the full-size picture shown in the last update gives a pretty good idea of what I intend.
Hi Pete, I can appreciate the amount of work you have put in to these ejectors and they look really good....you may wish to note that on the full size items, the larger hexagon on the ejector body is actually "cast in" and is threaded internally to suit a thread put on the outside of the sand pipe. The hexagon is for a spanner to screw the ejector on to the pipe end.
I know you've made them now etc. etc. but I just thought I'd mention it!
Hi Pete, I can appreciate the amount of work you have put in to these ejectors and they look really good....you may wish to note that on the full size items, the larger hexagon on the ejector body is actually "cast in" and is threaded internally to suit a thread put on the outside of the sand pipe. The hexagon is for a spanner to screw the ejector on to the pipe end.
I know you've made them now etc. etc. but I just thought I'd mention it!
Cheers Don
Hi Don
Think I understand what your saying...that doesn't matter for me, the end result looks the same or close as...
Good evening chaps..ok so a slight change of plan... after I posted the last update it was pointed out to me by Kevin Fisher and Peter Pope (FB) who I believe both worked on 4472, that I had copied the wrong ejectors, evidently the ones in my photo from 2016 are air ejectors off an 08 shunter, now this would never do. First I must say a big thank you to these two gentlemen for pointing out my error for which I'm most grateful. having learnt this I duly took a look through my photo's from the 30's to see what I (yes I know, I should have done this to begin with) could find. I found 3 pictures where once I zoomed in I could get a good idea of what the ejectors really looked like, as it happens Don's drawing is very good although he shows no union connections in his drawing. To be fair when looking at the photos I had the ejectors and connecting pipes all looked different and one early photo had no nut for the sand pipe connection but did for the steam pipe. I eventually settled on a photo which was taken in 1934 after 4472 broke the 100mph barrier, the photo in question showed the driver and firemen standing in front of the coupled wheels where the steam ejector was clearly visible. It looks like it had union/nuts for both feeds but I won't swear to it as it is a little blurred when zooming in. Doesn't matter to me as I said before I chose to add the connection points for ease of maintenance.
Here's the photo in question...Note that there's no flange or hanging bracket and that the orientation of the pipes is different, with the sand pipe angled directly above the ejector rather than to the inside as seen on the air ejectors. It's also very clear to see how much longer the steam feed is before meeting the sand feed and also that it's now in line with the wheel flange rather than being angled inboard resulting in the sand pipe being more in line with the wheel flange.. So I needed to make a decision, do I remake them, or do I modify those already made. After taking a look at what was involved I chose the latter, well I do so enjoy a bit of grinding.....not.
After spending most of the day working on these the result os shown here. I have extended the steam inlet, I could have made this a permanent attachment but decided to leave it as a screw-on part in case I ever get a stuck jet where it will be much easier to remove the extension and grip the protruding jet than trying to tap it through from the front. The bracket has gone although the part around the nozzle remains to give that part more length to match the photo, the flange wings were cut off and then all blended in removing much of the extra width of the sand feed body. I then soldered on a small washer to give me a finished front which was backfilled with solder which again was blended in. Also, I have reshaped much of the underside of the sand feed body, again to better match the prototype. It's not perfect and I could define a crease between the two body parts but this might exaggerate that the two parts are of differing sizes for the reasons given, it's a damn sight closer than the air ejectors were, ie they look like steam ejectors now. Once I'm happy with both ejectors they will now be painted black, with the nuts left bare metal, again in an attempt to match the prototype. This all took a good few hours today but I'm happy with the end result and again wish to thank both Kevin Fisher and Peter Pope for their input.
To finish with I decided to test the ejectors using air, the compressor was set at 60/65PSI, any more resulted in the temporary pipes getting blown off which I couldn't physicaly hold as I needed my other hand to operate the camera so used a lower setting. I'm happy with the test, it's an old boost gauge but looks like I was getting 6-7 inHG. BTW the jet is a 1" length of 1/16 copper pipe with one end swaged so that it doesn't get blown through. I have followed Don's drawing as to where the jet ends in the nozzle, I guess I could play around with this to see if the result changes... perhaps that's for another day
Thanks for looking in all, next update should see the steam sanders finished and then I'll be on to the gravity sanders...
I used to work with Kevin on the Asset Inspection Train project at Northfields depot. He did indeed work on 4472 at Southall depot during the McAlpine and Marchington ownership years, under the direction of Roland Kennington.
Lucky he spotted the modern change, but I think you've made a good job of the replacements.
Thanks, Dan.. yes IIRC he told me that he fitted the air ejectors... I have so many folk on facebook who kindly share their experiences with the full-size loco, it's a godsend...
I've been a bit busy with grandparent duties since the last update but have this morning done a little more on the steam sand ejectors. I first needed to check that the other ejector was working and once happy with that I played around with the nozzle jets. I have now shortened them a little so that as you look at the ejector in the first picture image the jet is just a little forward of the main body into the nozzle itself. This has increased the vacuum by about 30% and still only at 65 PSI, I would expect the inHG to increase further at the max boiler operating pressure of 100 PSI. Only two pictures for today, the first shows one of the ejectors in its correct orientation/position in front of the main driving wheel. The steam pipe is just temporary to test that the sandbox works, it will be routed in a slightly different way once painted/finished, I can see from the works drawings that it goes up to the stay above and joins a 'tee' piece, there's a lot going on in the drawings which may take some time to plot its proper route and what I don't need to fit. In simple terms, I need to meet the pipe from the other side and then route it back to the manifold turret which I obviously can't do for while yet. I think what I will do, is decide on where the two pipes join the 'tee' and fit those accordingly, the section to the cab will have to wait for now.
I have tried to take a video of the steam sandbox working using air, it's very difficult to film the sand leaving the nozzle as it's so small and faint in colour, I have placed a sheet of W&D under the wheel so you can see some of the sand as it builds up, most went on the floor... I have directed the nozzle a little outboard of the wheel, one to help show the sand but also to stop sand from getting blown too much around the model as many parts have oil on and I don't want the sand to mix in with this.
Tomorrow I'll finish the ejectors and pipework and paint the parts that need painting black. I'm in two minds as to whether to fit the sandboxes properly to the frames, I'm worried about knocking them but then I can't really fit the inside connecting rod until these parts are fitted so I guess my hands are tied, I'll just have to be careful. Next up will be to finish the gravity sanders and get these working, once that's done I can fit for the last time the outside motion brackets and then hopefully, get back on with the rest of the moving parts. During this, though I need to repaint the smokebox and mount that for the last time too as this needs to be done before I can fit the outside cylinders and their slide bars... I tell you the list of parts is endless and they all need doing before each other, or so it currently seems...thanks for looking in guys...
Just to clarify something what I wrote in my last update....I said that the steam supply goes to the manifold, well it does but what I should have said is that it terminates there, with the isolator valve, there is, of course, the steam valve in between. IIRC this is attached to the reverser but I'll look closer at this when the time comes...