NSW 'B' class mogul

[David]

After posting the steps I figured I had nothing better to do so I went onto another job I keep forgetting until I see Barry's loco which is a pair of small strips that go at the bottom of the smokebox wrapper.

Making them was uneventful enough and drilling the first side went ok but I blunted or broke all my 1.6mm HSS drills doing the second side. It's like the steel has hard spots in it - it drills easily in some places and not at all in others. I thought it might be near welds but there are no welds near most of the holes that wouldn't go.

After making the strips I scribed a center line on them and taped them to a piece of MDF using its straight edge as a reference. I eyed the drill tip against the edge of the strip and the center line to set the zeros. I held the strip while drilling - the tape was more to keep it lined up.



I used a couple of clamps to hold it in place while spotting the holes through the smokebox. The holes go all the way through the wrapper and the 5mm steel.



Then a couple of chamfers filed on the ends to finish up. These are not the final rivets, they're just there to see how it looks. I'd like to rivet at least a few over to keep the strips down but not sure how I'll do it because there's not much room in there.

[mbrown]

Have you tried a rivet squeezer?

Malcolm

[David]

No Malcolm, I don't have one and don't think anyone I know has one either. It would be a useful device for sure.

I'm always put off making riveting tools because I can't think how to make the dimples so the rivet head doesn't deform. I assume a ball nose cutter is involved, but what radius?

[Roger]

Dec 3, 2019 23:00:47 GMT David said:

No Malcolm, I don't have one and don't think anyone I know has one either. It would be a useful device for sure.

I'm always put off making riveting tools because I can't think how to make the dimples so the rivet head doesn't deform. I assume a ball nose cutter is involved, but what radius?

Yes, that's what I use. I just guess the size and then see how nicely the head fits. You can have a pretty good estimate if you hold a drill against the rivet head to see if the curve is similar if you don't have radius gauges.
If you make a press that has say a 3 or 4mm hole in it, you can use Silver Steel of that size to make short anvils of whatever size you need. You can add flats to them and alter how deep the pocket is to suit what you're doing. It's well worth taking a bit of time to design and make something that will work in a wide range of places.

[Jim]

Dec 3, 2019 23:00:47 GMT David said:

No Malcolm, I don't have one and don't think anyone I know has one either. It would be a useful device for sure.

I'm always put off making riveting tools because I can't think how to make the dimples so the rivet head doesn't deform. I assume a ball nose cutter is involved, but what radius?

On the basis of keeping it simple I bought two sets of hollow nosed nail punches, one to hold the rivet head, the other to peen the tail over. All the rivetting on the Britannia was done with these nail punches. The set of 4 pinches covers the range of rivet heads we use.

Hope this helps.

Jim

[David]

Slowed down a bit the last few days but still made the clack bodies and what I assume are inspection/maintenance covers for the sandboxes.



I soldered the covers to a scrap of brass for spotting the bolt holes. This made it easier to hold it and use the centre finder on it.

[Jim]

Very nice work David you should be very pleased with that. I certainly would be.

Jim.

[David]

Thanks Jim.

More sandbox details, done on the CNC machine this time simply because I can't file a hex to save my life. The hexes are 1.5mm AF and 0.5mm deep. The blanks were turned to just over size on the lathe then a 6mm endmill brought them to diameter and put the step in, and a 2.5mm endmill did the hexes.

They were held by their stems in a 5C collet in a square block. One of them is sitting in the washer that was used to give a bit of machining space above the collet.

[Jim]

Looking very good indeed David.

Like you I gave up on small hexs and used 12 and 16 BA hex head screw suitably shortened and held in place with a ouch of solder.

Jim

[David]

Where do you get those small screws? I think I bought some 16BA from Eileen's Emporium for laughs. Post-B-class I think I'd like to use those German screws and go metric.

[Jim]

I got mine from the model railway suppliers such as Woodpecker Models. Worth googling to find them.

Jim.

[David]

Bit of a disappointment today, although I'd say the part is usable as is. I don't have a ball-nosed endmill larger than 2mm dia, so decided to pocket out the sandbox filler lids on the CNC.

This is how I modelled the rounded cut-out in the lid. The cutting tool was a cylinder with each end chamfered to a ball. It doesn't look faceted to me.





At about the 3 hour mark it was looking pretty good. At the end, after nearly 5 1/2 hours these facets have appeared. They were not there when I'd checked earlier so they must be the end of the toolpath.



I tried a pocket toolpath to see if it would clean them up but it only touched the edges to around 20% of the way down.



I just simulated a pocket toolpath with 'rest machining' turned off and when I look closely at the simulation output the facets are there. So they must be part of the model.

I'm sure if I put some extra paint down there it will smooth out. I'm annoyed it looked better at the half-way mark! I don't think I'll hang around trying to catch it at the right spot or changing the program to try and get both sides at the 'good' point before faceting either of them.

One bonus was because I'm holding these in the square collet block like the last little parts the machine X/Y zero was still correct.

While this was being gently scraped out I was bringing some brass down to size to make the parts on top of the splashers that on the real one must give clearance to the coupling rod oil pots.

[Roger]

Hi David,
I had a similar problem on Alibre. It turned out that there is a setting in the Properties of the job under 'Display' that shows 'Render Quality' and as slider for 'Facet resolution'. Although that might seem irrelevant to the CAM output, it's clear that it uses this not only for the display but for the CAM output also. I suppose that's logical, but far from obvious.

It may well be that Fusion 360 does the same. If you zoom in on the model and take a close look, you can do the same if you can then find the relevant display property to see if it looks any different.

[cplmickey]

Nov 9, 2018 23:26:51 GMT David said:

The last job was to make little brackets to keep the operating rod in line and stop it flexing too much. I spent a few days(!) thinking about this and tried the idea I came up with and it worked a treat. The pictures should describe it well enough. The only important dimension is the inside radius of the bracket which I made 4mm to give some clearance around the 1/8" operating rod. So a 6mm hole in the former and a 4mm drill shank allow the bend to be made in 1mm steel.

The other trick is to form the bend with the drill shank and then just squash the bracket further without the drill shank to clean up the corners.

That's really neat. I've got a few of these to make and this seems an excellent way of doing it. Ian

[David]

With a sore everything, hot weather, and a lot of work to do I wasn't in the mood to do much on the loco this week but as Barry mentioned in his thread a break of a week can turn into a break of a year. I've had six month breaks before on this project so at least did my designated Monday night sessions that my wife insisted I commit to do to ensure something gets done each week. It's easy enough now in the summer but I don't think I'll be so keen in winter.

So last week was the guard irons and tonight saw a couple of gussets that support the front buffers. I managed to bend both guard irons exactly the wrong way but by the time I was sorting it out I was getting pretty relaxed about the bending angles etc so it didn't take long to fix. The gussets proved I learned nothing about bending metal doing the steps a few weeks ago because the bent up bits of the gussets also turned out 1mm too high. That's the thickness of the material. I didn't allow any extra this time so no idea how to calculate these things.

The parts are mundane and the workmanship is nothing to talk about but it's 4 pieces that didn't exist 8 days ago. I wish I'd used one size smaller screw on the guard irons but you can't easily see the screw heads under the gussets and running boards.

I picked up some hollow headed nail punches from a junk shop today so I can look at using them in a rivet squeezer sometime. They were $3 AUD for 2 of different sizes so no great loss if I mess it up. I'm assuming at that price they're not very hard but should be fine brass rivets.

[David]

I'm still wanting to paint the frames etc before I put the wheels back in and make the final push on the upper works so no big jobs happening at present. But I've done a few small things.

• Put some loctite on the die block pins and the bolts that secure the expansion links to the eccentric rods. One die block pin was constantly unscrewing itself when running on air.
• Took the mill scale off the guard irons and softened the edges a bit.
• Took the mill scale off the stretcher that goes beneath the steam chests. I used the phosphoric acid to do this, leaving the part in for about 15 minutes. I think it could have done with a bit more but I didn't want to risk ruining it. I lost a tender handrail bracket to the acid years ago when I forgot about it so I've been wary of it since for use on steel.
• Attached the cab steps. I had to take 3mm off the ends of the brass angle the footplates sit on to get the steps in so they'd not stick out beyond the footplates. I don't know where the real ones were in this regard but followed the advice it was best to keep them tucked in so they didn't catch on things.

The cab steps were easy enough except on the first one I put the holes for the bolts securing them to the foot plate too close to the angle's web. I do this a lot, despite always wondering if they'll be too close and then thinking 'she'll be right'. They were so close I couldn't get a hex headed bolt to turn in one of them - never mind get a nut runner over it - but just as I was about to fill in the holes in the steps and footplate I thought I'd just turn the hex heads round and put a slot in them. Ugly, but a lot less work and you can't see it. This can be explained in part by the fact I was going to use small headed 8BA screws before deciding to go with the instructions and use 6BA ones. But they were still stupidly close.





It all looks a bit rough in that photo but I think it looks better in real life.

This year started quite slowly with only a reversing screw and reach rod, and boiler and firebox cladding to show for about the first 7 or 8 months. But it's picked up a bit since then thank goodness and looks quite a lot like a loco when it's all together.

[David]

A long day to achieve very little! I couldn't get back to sleep after I got woken up so I fired up the CAD at about 4:45am and tried to model the little curved valance that goes at the front and back of the cab roof, thinking I'd cut them on the CNC machine. After about 2 hours of that I gave up. The cab roof is a compound curve and even though I did the CAD drawings for the cab parts it's difficult to extend a compound curve for the overhang at each side. I'm going to leave it be until I have the roof rolled and see if that helps me.

So then I started on a drawhook as I need one for the front buffer beam. Again, no luck trying to draw it to the design even though I've already made one by hand for the tender. I knew I had a drawing of the prototype one somewhere so I found that, scanned it, scaled it, and modelled over it. It's not the same as the one on the tender but it will do!

Then a couple of hours programming it - I wasn't at my best. Mid afternoon I started cutting it out. I tried the method of tool length offset setting Wilf described to me - I used the shank of a broken 20mm endmill as a gauge and first brought the empty spindle down on it to set a zero point then passed it under each tool in turn to get their lengths. Starting with the tool not being able to go under and moving the spindle up until it could. This was pretty easy.

Then I managed to ram the spotting drill into the stock when trying to set the zero height for the job - the jog was on 'continual' rather than 'step'. Didn't break the drill luckily but it couldn't have been good for anything involved and left a mark on the job. I thought I was outside the part when I did this but I wasn't. So that was all a big cock-up.

I meant to stop cutting the profile 0.2mm above full depth to keep the drawhook attached to the stock. On the last pass the cutter went right through the stock. Luckily the end of the cut on the curved part of the hook had just a sliver to material left and I let it go until it looked/sounded like it was going to let go and pressed the stop button I was hovering over. I got away with it.

I measured the stock afterwards and even though it's marked 6x90mm from the shop it's only 5.7mm thick. That explains the cut-through.


This photo shows I'll need to make the shank shorter so it doesn't foul the stretcher. That's a good thing because I still need to make the end of the shank circular to cut an M6 thread on it and I think the shorter the shank the better that will go. I'm considering cutting it where it should turn round and silver soldering a pre-threaded stud on, especially now I know the stock is only 5.7mm thick to begin with.



That was a silly amount of time and effort for a drawhook. But I wouldn't have made one by hand today and I have most of one done now so it's a net win. I tried a new method of setting tool lengths so that was another bonus.

[Jim]

Looks great David and well worth the effort. Well done.

Jim

[Roger]

Dec 28, 2019 10:00:16 GMT David said:

A long day to achieve very little! I couldn't get back to sleep after I got woken up so I fired up the CAD at about 4:45am and tried to model the little curved valance that goes at the front and back of the cab roof, thinking I'd cut them on the CNC machine. After about 2 hours of that I gave up. The cab roof is a compound curve and even though I did the CAD drawings for the cab parts it's difficult to extend a compound curve for the overhang at each side. I'm going to leave it be until I have the roof rolled and see if that helps me.

So then I started on a drawhook as I need one for the front buffer beam. Again, no luck trying to draw it to the design even though I've already made one by hand for the tender. I knew I had a drawing of the prototype one somewhere so I found that, scanned it, scaled it, and modelled over it. It's not the same as the one on the tender but it will do!

Then a couple of hours programming it - I wasn't at my best. Mid afternoon I started cutting it out. I tried the method of tool length offset setting Wilf described to me - I used the shank of a broken 20mm endmill as a gauge and first brought the empty spindle down on it to set a zero point then passed it under each tool in turn to get their lengths. Starting with the tool not being able to go under and moving the spindle up until it could. This was pretty easy.

Then I managed to ram the spotting drill into the stock when trying to set the zero height for the job - the jog was on 'continual' rather than 'step'. Didn't break the drill luckily but it couldn't have been good for anything involved and left a mark on the job. I thought I was outside the part when I did this but I wasn't. So that was all a big cock-up.

I meant to stop cutting the profile 0.2mm above full depth to keep the drawhook attached to the stock. On the last pass the cutter went right through the stock. Luckily the end of the cut on the curved part of the hook had just a sliver to material left and I let it go until it looked/sounded like it was going to let go and pressed the stop button I was hovering over. I got away with it.

I measured the stock afterwards and even though it's marked 6x90mm from the shop it's only 5.7mm thick. That explains the cut-through.


This photo shows I'll need to make the shank shorter so it doesn't foul the stretcher. That's a good thing because I still need to make the end of the shank circular to cut an M6 thread on it and I think the shorter the shank the better that will go. I'm considering cutting it where it should turn round and silver soldering a pre-threaded stud on, especially now I know the stock is only 5.7mm thick to begin with.



That was a silly amount of time and effort for a drawhook. But I wouldn't have made one by hand today and I have most of one done now so it's a net win. I tried a new method of setting tool lengths so that was another bonus.

Hi David,
That's a good result, and a lucky escape. I'm afraid any stock plate is always going to be under size, some by quite a lot. I always measure what I've got. It's worth leaving nibs as well as the wafer if you're going to use a wafer at all. These days I go to the exact depth and use nibs alone. I make them very long, say 12-15mm and no less than 0.5mm high. That gives plenty of support and also a very gentle slope up and down.

[jasonb]

What are you using to do the programming? Couple of hours seems an awful long time for a profile and one possibly two holes.

How is your Dome doing?