NSW 'B' class mogul

[David]

Roger, F360 used to handle the rapid moves - it's not something you had a choice with except the Z ones I guess you could set one of the heights to a long way up that the rapid moved to and then it would use the feed rate for the long way down. I don't think my reactions would ever be quick enough to save ploughing into a clamp or vise jaw whatever happens. I've never managed it so far!

I spent the afternoon and evening working on these brackets and brass standoffs. It all started well enough but after I made a paper mock up and then the initial bent bracket without drama only to find it didn't sit perpendicular to the running boards it all pretty much went to hell for the next 8 hours. I had to straighten it, bend it about 3mm further down, plug the bottom screw holes I'd already drilled which just happened to butt up against both the new bend AND the new hole locations, ... it's now bent, twisted, and dented every which way.

The brass standoffs didn't go without incident either. With all the luck I'd been having on the CNC I decided to use that and the first thing I did was snap my virtually new carbide 6mm endmill when I was using it to find the edges (great precision was not necessary) when I forgot to switch from continuous to stepped jogging after moving the cutter near the edge. I guess I have good 6mm dowel now.

Then I decided to mark out, centre punch, and drill some holes the usual way. And spent a lot of time elongating them with a file. I did the rest using coordinate drilling on the manual mill.

Glad to have this part just about done aside from some cutting screws and the lower plate to length.

Here's the top part of the bracket pre-adjustment. It looks a lot more beat up now.



Here it's all together. The bottom plate needs to be made shorter. It's a real puzzle putting it all together - which screws go in first takes a few times to figure it out.

[David]

Added a few bits to the steam cylinder and got it mounted on the smokebox. I have to make a flange for the elbow so it can screw onto the smokebox. It took a lot of time and a couple of goes to make the elbow!

On one of the photos above you can see the through hole was started with a 6mm end mill, whereas the hole is only 5.5mm. I'm sure I had some 5.5mm endmills from various batches of cheap HSS endmills via ebay but couldn't find one. Anyway that step has been filled with soft solder so can't be seen.

The end caps of the shuttle valve housing and the governor casting will be loctited in place and the elbow might just be held in place via the screws to the smokebox - not sure about that yet.

[Jim]

You have done a great job there David in creating a nicely detailed pump that was such a feature of NSWGR locomotives.

Jim

[David]

Thanks Jim. Bit of an oddball pump I think. Barry said it's the smallest one NSWGR used. No cooling ribs, no visible forked pipes on the air cylinder, it must have been a real baby. But with his drawings and photos and the ones I got from the Powerhouse museum it's going well. Still lots of piping etc to go, and a few bits missing from the cylinders. I'm flagging a bit now, after about 3 weeks on it! There are a lot of simplifications where cast features are not easily replicated on the shuttle valve housing and the various cylindrical bits that are above and below the air cylinder. The only things that bug me are missing bolt heads (too hard), I think the little cylinders on top of the air cylinder should be smaller and closer to the spacer but I couldn't seem to do that even in the 3D model, and those missing cast shapes that would just take too much time and beyond my ability to solder them all on without the whole thing falling to bits!

When I saw Barry's one my first thought was "that looks good" so if mine gets the same response I'll be satisfied. I'm lucky in that I have very little idea what a real NSW steam loco looks like so this is one of the first times I've really noticed what's missing and only then after I remembered I had photos of one.

[Jim]

If it's any help David I made up these dummy hex headed studs from small brass dome head rivets.

I used a small chuck on an hexagonal shaft which was then mounted in the vice to hold the rivet while I gently filed the flat for the nut.

a light cut in the lathe created the end of the stud just showing from the nut. It's a bit painstaking but I got what I wanted.

Jim

[David]

Very nice! But beyond my skill with my hands. I can't file evenly for nuts. I've tried to file small hexes numerous times and thrown them away every time.

The bolts I'd have to make could be made from standard 10BA bolts. I don't know how to describe the shape other than to say it's like a semicircle the width of the threaded shank. This must have been because of the small width of the flange they were sitting on - it wasn't wide enough for a full hex to sit on. So it would only involve filing the existing head down and rounding it over but I wouldn't be able to make 2 the same, and wouldn't want to make all of them!

[jasonb]

Should be able to set that up for easy cutting in the CNC, just one profile to produce the exposed stud and another to do the hex. Hold part vwrtically in a collet block or simple block

[jasonb]

Even those half moon bolts would be easy on the CNC. Simple fixture to hold bolt horizontal then mill one side of the existing hex head flat, reposition fixture 180deg to machine the other flat followed by the curve.

I have not shown the part being flipped over but cut from underside would be the first cut. Also show the existing head as square rather than hex but could also have made the bolt from round stock. Note on the profile tool is set to stay down and cut in both directions to reduce amount of rapids.

[David]

Dec 1, 2020 7:33:10 GMT jasonb said:

Should be able to set that up for easy cutting in the CNC, just one profile to produce the exposed stud and another to do the hex. Hold part vwrtically in a collet block or simple block

The studs and hexes were done on the shuttle valve covers with the CNC. Couldn't be bothered with the half-moon bolts and I'm not sure I've left enough room for them - the holes may come out too close to the cylinder wall.

[David]

The pump is nearly complete. I think there are only 2 machined parts left which are the cap and bolt that go at the top of the steam cylinder and the flange where the exhaust goes into the smokebox. Other than that it's all pipes, which I don't have the wire for. I have revised the model to include an extra bolt hole on the front top and bottom of the air cylinder. The photos show 5 bolts here and I only have 4.



Then it was back to the cab roof. I knew this would be a difficult, being bent sheet metal. The roof was bent by a friend about a month ago but I knew it didn't fit well for various reasons - I was too dumb to put a constant radius curve on the spectacle plate parts and when they're attached to the loco they don't line up quite right. I decided it would be easier to make the spectacle plate conform to the roof than going the other way. So a few hours with a carbide burr, sanding drums, and a file got a good enough fit. I was taking it very slowly with a lot of checking.

The increasing curve to the edges and the left/right mismatch.



When the roof was sitting on the cab sides the spectacle plate was above it for almost the whole width so at least 1.6mm had to come off everywhere except the very ends.



Here it is post-fitting with the angles that will allow it to be screwed to the cab sides. Those angles were a right pain! I had to cut them down from 3/8 to 1/4 and by that time I'd been going all day and took the dumb decision to use the mill as I usually do. That just resulted in cutting into a parallel which should have had just enough clearance but vise jaw lift defeated me. At which point I thought "why not just use the bandsaw?" Then they had to be opened a bit because the angle between the cab side and roof is greater than 90deg. This went ok for one and terribly for the other which required a lot of faffing around to get it back close to straight, although it's still pretty bad. I don't care how it looks but I don't want it to distort anything when the screws go in.



I wanted to get the angles on and done, but then stopped because once I start making mistakes they just get worse as the day goes on. This was over 6 hours for me as it was.

[David]

Small progress.

The exhaust pipe and flange for the air pump. The flange should be closer to the bracket but it just ended up where the pipe allowed it to go. I doubt anyone will comment.



I've made a start on the whistle. I tried to buy a casting but haven't been able to get one. Luckily the father-in-law has a full size one in his workshop, and let me borrow a spare casting set too.







The rough edges on the whistle is my attempt to round over the corners to make it more casting-like. The vanes separating the whistle chambers don't have the correct profile but it's good enough. Still a few fiddly bits to make that this assembly sits on.

The new 4mm 3 flute carbide endmill tore through all this beautifully. I've had it cutting at up to 120mm/min, but that was just the tiny flange above. The whistle bits were done at 90mm/min and there was no protesting, I could barely hear it cutting over the spindle noise.

[David]

I've held off posting any more about the whistle until it was finished, which it now is.

Here are most of the parts, except the tricky clevis the lever goes in.



The parts on the left are the casting kit you can buy, when it's in stock. On the right are mine. At this point the shaft has been silver soldered into the body and the bottom nut secures the sandwich of all the other parts but squeezing them up against the body.



It took me a long time to figure out how I was going to make the clevis. Step 1, done on the CNC mill.



Step 2, removing all the stock left around it on the manual mill so I can cut the slot.



Step 3, cutting the slot. The bar had to be brought a long way out of the collet to give clearance to the tool collet over the stock collet block.



Step 4, part off the part on a stick.



Step 5, the cleaned up clevis and a 6BA nut that forms the other part of this piece.



Step 6, nut and clevis silver soldered together and the dummy plunger the lever acts upon.



The finished item with the clevis/nut soldered onto the cross shaped body part:



Here's the real thing:



That took quite a long time!

[Jim]

Superb craftsmanship David. Well done!

Jim.

[barlowworks]

That looks amazing, well done. 👍😎

Mike

[mugbuilder]

Loverly job Dave.

[David]

The goal was simple enough - make and fit the regulator handle. But I've spent most of the last couple of days trying to dig myself out of numerous traps I set for myself before this could be done.

The first problem was the spindle for the valve in the lower gauge glass fitting was too long by about 10mm and both looked bad and would get in the way of the regulator handle. This spindle is stainless with 5 and 8 BA threads and captive in the fitting. I thought the fitting had to come out to remove the spindle - it doesn't. But I didn't think of that at the time, so to get the fitting out I had to remove one of the knee guards from the dummy backhead. I tried to undo the 10BA screws holding it in and they both sheared off. All the steel screws holding the detail onto the dummy backhead had rusted in. There must have been enough bakers fluid hiding in the wrapper join to work it's magic over the last x months. That stuff is evil.

I soaked the rest of the screws with WD40 overnight, and it worked for 2 of them. I gave the other two another night, but they still snapped off. I used a 1.2mm carbide PCB drill in the CNC machine to drill the holes out. I used the CNC machine because it has the highest spindle speed and I could control the downfeed with the jogging controls much better than I could on the manual mill. I didn't manage to get the drill in the middle of the screws, but the holes are retapped and I'll have to make the clearance holes in the detail parts bigger to handle the offsets.

Then while retapping the holes I managed to pop the firehole door ring off with my fingers. Whatever JB Weld is good for, a brass to brass joint isn't it. I'm stumped on how to attach this part - I've failed to soft solder it twice, and glue it with JB weld twice. I don't have enough room to rivet it. I've made two of them and may have to make a 3rd with a larger flange. I don't want to, they waste a lot of expensive large dia brass bar.

I'm right sick of that bl**dy dummy backhead.

To shorten the valve spindle I made a split 5BA collet to hold it, cut some off the end, and turned some of the existing 5BA thread down to 8BA size. Trying to use a die on this didn't go well, but I got enough of a thread on that I could force a brass nut down it. They are now a matched pair. I haven't made the handle yet.

Next I made the boss for the regulator handle with a 3.6mm square hole in it. I used the CNC machine to do the square but could only get down 5 of the 6mm required so finished it with a file.

With that done I looked at Barry's photos of his two cabs to see where the regulator handle was and marked the spindle for where the square and thread had to go. Then I realised that the spindle is captured in the regulator - I wanted to call the regulator "done" a year or two ago so soldered it all up. Now I had a turning operation to do with the whole of the regulator in the lathe spindle and sticking out the end. I just stuffed a rag around the end of it and didn't run the lathe too fast. At least it took a 4BA thread relatively easily, despite also being stainless.

Using the material I'd cut off the end to get the regulator spindle to length, I cut a test square on it with the manual mill and got it to be a good fit. I left the cutter locked at that height to mill the square on the real one. When I tried the boss on it, it was extremely loose. I measured the square and it was 0.4mm smaller in both directions than the test cut in the same material at the same settings. How does that happen? It was only a 0.5mm cut on each side with the cut very close to the collet in a square collet block so deflection shouldn't have been a problem.

I tightened the nut up against it and it stopped moving so I'm going to live with it until it proves to be a problem.

After all that I was too hacked off to want to make the rest of the handle.

[jon38r80]

A "just job" then, as in 'I will just do this quick and simple job', make a right mess of it and expand the job into an epic saga   

[David]

LOL, that's it!

Given my inability to attach a simple top hat shaped fire hole ring to the dummy backhead I have come up with the idea of making one that can use the screw holes from the door and knee guards to be attached. One part cut from sheet with lugs that go to the various screw holes, with the existing ring trimmed and silver soldered to it. The lugs would be mostly hidden, but there is no effort to make the backhead or cab fittings scale in any case so it won't matter that it looks a bit odd.

The next time I go out to the workshop it will be to finish the regulator handle, and finally make the little cap for the top of the air pump steam cylinder. That should be trivial, and hopefully not another "just" job.

I also figured I can put a grub screw into the regulator handle boss if necessary. If that bears against one of the flats on the square it will tighten things up. But it will make it obvious I've stuffed up the hub/spindle area and the square was a waste of time.

[David]

I was going to make a new boss this morning and got as far as changing the CAD model and generating the g-code before I realised the square would be smaller than the OD of the thread so the boss wouldn't go over the it!

So I made an L shaped shim from some thin brass boiler band offcut. The first one I made seemed sloppy after the solder operations so I made another and used the nut to drive the boss over it. It will be interesting trying to get the handle off again!

Then I made the rest of the handle and other than it being quite tricky to solder together it went without incident. I had to redo the boss joint after it moved and put the lever off center, and it could stand to be done again to get a better fillet but I'm going to leave it as good enough.

[Jim]

You're doing a great job there David and as for the boss on the end of the regulator lever I think it looks just right. Having been down that path I've discovered that redoing things unless they are clearly wrong are best left as is to avoid a total disaster. I have more than my share of disasters that happened when I decided to 'just redo that bit'

Jim