NSW 'B' class mogul

[Jim]

Great to meet up with you today David and see your NSWGR B Class in the flesh. It's going to be a lovely model when completed.

The club's NSW Scale Day is a great opportunity to see some superb scale examples of NSW locomotives and rolling stock either in the steaming bays getting ready for a run or out on the track hauling mostly typical goods trains.

Ben De Gabriel's 50 class.

The steaming bays with a nice 59 class getting ready.

All in all it was great to catch up with everyone and to come away with enthusiasm rekindled to haul the Britannia over for a steam up.

Jim

[David]

It was great to meet you too Jim, and nice to have the Canberra division of the club up for a visit! My son was quite pleased with the long train formed at the end of the day when most of the wagons present were put together and triple headed by the 50s. I'm looking forward to both seeing your loco and watching progress on the beam engine.

I made a start on the firebox cladding yesterday. I'll say no more about it until it's done. I may be some time.

[David]

I've spent the week trying to get the RapidTurn going and make a safety valve body. The first one I made was pretty good, but the model had a few bits I didn't like. So I tweaked it and kept fiddling with the program until I could get the whole pillar turned, threaded, and parted off.

I'm still struggling with repeatability on the RapidTurn and the threads are too loose for my liking but I doubt I could turn one of these to an acceptable standard never mind get two to look anything like the same so I think these are good enough. Unless I manage to make some better ones.

Today was spent making the fixture to hold them in the vice for drilling and boring. The two safety valves are different enough on their tapered diameter that one of them sits a few mm higher in the taper than 'designed' and the other one bottoms out! On the other side the 20mm bore the flange sits in to drill the bolt holes doesn't fit either of them particularly well, and one valve flange is quite oversized. I can always turn it down on the manual lathe but it's irritating they so different. I think I can fix the bottoming out in the taper by shaving 1mm off the top of the fixture on that side. I'll make some more bodies and see which situation is more common before I do it.

The bad finish in the straight bore is due to the shank of the 6mm endmill rubbing. The flutes are only 15mm long and the bore is 25mm.

[Jim]

They look superb David and so typical of the class.

Cheers

Jim

[David]

Thanks Jim.

Tonight was spent trying the fixture out for drilling and boring the bodies of the safety valve. I'm drilling through 5.5mm, then drilling a bit under 8mm nearly down to the flat seat depth. Then an 8mm endmill bores the hole to 8.8mm for tapping 3/8 x 32. If I had a 55deg thread mill I'd then try and mill the thread. But I don't so that will be a manual operation on the lathe with a tap.

It worked well and held the pillar securely, even the one that looked like it was bottoming out. The only shortcoming is I had planned to butt the machined edge even with the vise jaw ends so I didn't have to indicate every part. But the pillar sticks out the bottom too far and either hits the slideway on the vise or the drill will go into it at least. I have enough aluminium to make another, longer fixture if necessary where the bore will be over the centre of the vise. Given this fixture was an experiment it's a bonus it works at all.

In the next couple of days I'll try out the other side of the fixture where I'll drill the 6 bolt holes in the flange. That's going to be more of a challenge as the flange diameter varies over about 0.1mm and the fixture is probably bored fractionally undersized. I may have to clamp the part in the fixture as best I can, indicate in on the threaded end and hope that gives a good enough centre to the PCD so the studs and nuts don't look off.

I'm itching to get one of the Impact Tolerant Touch Probes but I spent that money on measuring equipment this morning. Some Moore & Wright calipers (2 of the valuelines to keep near the machines, 1 of the absolute for the bench) and a the cheapest Mitutoyo 0-25mm micrometer they had. Trying to figure out what's going on with the RapidTurn diameters with a pair of cheap, loose calipers is driving me mad. Machine DRO's easter sale looked like a good time to get some proper measuring stuff which I've always wanted but my arms were too short to reach my wallet. I have some inch micrometers I use when I'm working in that world but I don't know how accurate they are and I don't want to convert every time when working in metric.

[Jim]

Just in passing I have two digital calipers, one from Hare and Forbes $80 a few years old now and the other from Aldi at $20. The Aldi one is excellent with easy to see display and a push button to convert from metric to imperial.

Jim

[Roger]

Hi David,
Those look really good. I'm wondering why you're having repeatability issues. I'd want to completely finish the threaded end before even roughing the body so that the screw cutting is being done while there's the greatest amount of material to support it. I don't know what order you machined it in but that might be a source of the problem. If you do it that way, you have the opportunity to make the thread over size to start with and then home in on the fit you want before moving on the the rest of the body. If you go too far, you've only wasted a small amount of material.
When screw cutting, you can start with generous cuts for up to the first 2/3 of the thread depth, but then you need to finish with say two or three more to get really close and a single very fine pass to clean up and take the spring out of the tool. With a big overhang, you'll be surprised at how much it can deflect.

[David]

Jim, I'm using the Aldi ones too but they're getting old and loose. They are just as good as the more expensive ones you buy from H&F etc. Mine have a guessing range of about 0.04mm now.

Roger, I am cutting the thread as soon as possible, for exactly the reasons you give. I have the full 25mm dia of the bar right up to the undercut for the thread.

After I've done the stud holes in the flanges of the two bodies I have I'll go back and experiment with the RapidTurn some more. I'm going to lighten up all the cuts, and also dial back the thread depth and see if that gets better results. The cuts are not heavy as it is, but the whole dinky toolpost situation probably just isn't up to much. If I was really serious I'd have to make a solid gang tooling setup but I don't want to spend the time and money, as always. If I can do it with light cuts I'd prefer to do that.

I'm also trying to minimise the number of tool changes given their manual. That's another reason the grooving tool is doing so much of the work. I can do the facing and thread OD with a RH knife tool, then thread undercut with the grooving tool, then the threading tool, then everything else with the grooving tool. So 4 tool changes per part. Every tool change is wasted time and a chance for inaccuracy or other problems. Wouldn't matter if I was making 2 but I'd like to sell these.

I'm always reading carbide tooling doesn't work well with light cuts which is why I've tried to go with 'heavy' cuts - all of 0.5mm deep with 0.25mm finishes!

[David]

I tested my stud hole drilling program and it works. There was a near miss with the first spotting operation because I'd not set the Z height! Luckily the 4mm spot drill didn't break through either the edge or top of the flange so I got away with it - amazing luck it stopped before breaking through. I'm tired, cranky, and sick today so not ideal conditions for using automated machinery. That's my excuse anyway.

I programmed in the toolpaths to contour, thread, and part the caps. I double checked all my threading parameters and they are correct so I just upped the number of passes to 20(!). It's annoying I can't get F360 threads going. I have to do all the ops up to threading in one program, load in a threading program done from PathPilot conversational, then load another F360 generated program to do the rest. I should be able to get PathPilot to call out to a separate program via a 'manual NC' operation and an O code but haven't got it working yet. I could also just copy/paste the threading code into the F360 code but I'd lose it every time I regenerate things which I do quite a lot while tweaking. The caps are drilled in the manual lathe. The thread on the caps is only 3mm long with a 2mm wide undercut at the end, and the thread is beautiful - if nothing else I really love the CNC threading. If I get a smaller insert I can thread closer than 2mm. As it is the edge of the insert misses the stock by a whisker. I could do it by hand if I had a handle for the manual lathe spindle but I couldn't do it under power.

I also ran two more bodies off after reducing all the cutting depths on the toolpaths. I did them one straight after another with no adjustments to the X DRO. Dimensionally they're better - maybe .002 or .003 different on various diameters but the finish is getting worse on the tapered part. I think it would clean up fine if I used some scotchbrite or fine sandpaper while spinning them in the manual lathe.

So these are starting to come together. A little ridge is being left on the bottom of the flange due to the finishing toolpath not being just right. The bigger problems are getting the bodies and caps out of the 'nut' I hold them in on the manual lathe for tapping the 3/8 inside thread and drilling the caps. They get screwed in hard and I cannot undo them by hand. I need to make some soft collets like you use to get injector cones out so I don't mar them as I am doing now.

I also have no idea if the milled ball seat works and have no suitable balls or springs yet so I better get onto that.

[Roger]

Apr 20, 2019 0:06:34 GMT David said:

Jim, I'm using the Aldi ones too but they're getting old and loose. They are just as good as the more expensive ones you buy from H&F etc. Mine have a guessing range of about 0.04mm now.

Roger, I am cutting the thread as soon as possible, for exactly the reasons you give. I have the full 25mm dia of the bar right up to the undercut for the thread.

After I've done the stud holes in the flanges of the two bodies I have I'll go back and experiment with the RapidTurn some more. I'm going to lighten up all the cuts, and also dial back the thread depth and see if that gets better results. The cuts are not heavy as it is, but the whole dinky toolpost situation probably just isn't up to much. If I was really serious I'd have to make a solid gang tooling setup but I don't want to spend the time and money, as always. If I can do it with light cuts I'd prefer to do that.

I'm also trying to minimise the number of tool changes given their manual. That's another reason the grooving tool is doing so much of the work. I can do the facing and thread OD with a RH knife tool, then thread undercut with the grooving tool, then the threading tool, then everything else with the grooving tool. So 4 tool changes per part. Every tool change is wasted time and a chance for inaccuracy or other problems. Wouldn't matter if I was making 2 but I'd like to sell these.

I'm always reading carbide tooling doesn't work well with light cuts which is why I've tried to go with 'heavy' cuts - all of 0.5mm deep with 0.25mm finishes!

Hi David,
It sounds like you're doing all the right things. The difference between a sloppy and well fitting thread is a small one, so that's definitely gong to need a fine cut, say no more than 20 microns for the final cut.

Don't believe what non professionals say about carbide, most of it is utter nonsense. Carbide can be used with deep, fine, continuous or intermittent cuts and on any material you like, I do it all the time. Things like aluminium and plastics are better cut with polished razor sharp inserts though, sometimes I use those on Stainless and Phosphor Bronze too.

Don't take anyone's word for anything, try it yourself and see what works.

[David]

I now think the threads are ok. I double checked my start and end depths and they are correct and I increased the number of passes to 20. I think the problem is more with the boiler bushes. I ran a tap through the one that was tight and now it's loose so I think it was deformed during soldering and is now eccentric or something. Screwing the bodies into a tapped hole in brass and the fit of the caps into the bodies seems good enough.

No point going any further until I can test them. I've ordered some 440 stainless steel balls and have to find some springs. Google doesn't give me a consistent answer on the corrosion resistance of 440 stainless so I'll just have to try it. The balls from a local supplier were $3.00 each for 304 stainless plus postage. From Amazon I've got 100 for about $30 including postage. Given there are so many of them I hope they're ok!

I can't see that anything can go wrong with them except the ball seat isn't good enough. If that's the case I'll need to find either a longish 8mm endmill or make a D-bit. I've never had any luck with D-bits. It all looks simple enough but I haven't ever managed to make one work very well.

As for carbide I watched a video of a tour of Stefan Gotteswinter's workshop and he sharpens his carbide inserts razor sharp and he agrees with you that carbide can be just as sharp and take fine clean cuts. I think he uses a diamond wheel to sharpen the carbide.

[Roger]

Apr 21, 2019 7:38:23 GMT David said:

I now think the threads are ok. I double checked my start and end depths and they are correct and I increased the number of passes to 20. I think the problem is more with the boiler bushes. I ran a tap through the one that was tight and now it's loose so I think it was deformed during soldering and is now eccentric or something. Screwing the bodies into a tapped hole in brass and the fit of the caps into the bodies seems good enough.

No point going any further until I can test them. I've ordered some 440 stainless steel balls and have to find some springs. Google doesn't give me a consistent answer on the corrosion resistance of 440 stainless so I'll just have to try it. The balls from a local supplier were $3.00 each for 304 stainless plus postage. From Amazon I've got 100 for about $30 including postage. Given there are so many of them I hope they're ok!

I can't see that anything can go wrong with them except the ball seat isn't good enough. If that's the case I'll need to find either a longish 8mm endmill or make a D-bit. I've never had any luck with D-bits. It all looks simple enough but I haven't ever managed to make one work very well.

As for carbide I watched a video of a tour of Stefan Gotteswinter's workshop and he sharpens his carbide inserts razor sharp and he agrees with you that carbide can be just as sharp and take fine clean cuts. I think he uses a diamond wheel to sharpen the carbide.

Why not buy Silicon Nitride balls? They're cheap, harder than anything else and most grades have better spherical tolerances than those normally sold to Model Engineers. They won't deform if you press them into the seats and will not corrode.

[David]

Thanks Roger, I may try them. Do you think they might have bits chip off them? I told a few club members and one of them was wondering if the hardness could be a problem.

Finally enough progress to report on the cladding. I made the small plates that bolt to the frames and hold the bottom of the cladding in place so it finally sits about where it should. Took about 6 hours to make and fit 4 plates! The first one I made was too short so that added about an hour onto the process as I made another piece of 8mm x 1mm angle and got it into shape. I was working to the drawings whereas I should have realised that because I'm working with the boiler AND sheet metal, nothing was going to be to spec so should have gone from the job rather than the drawing. The cladding is about 10mm too long. But I'm not changing it now.

Next job is to make some threaded transfer punches to go in the safety valve bushes to mark the location of the holes for them. These will also be used in the clack bushes.

Then there is a ton of cleaning up to do on the brass part at the front, soldering it to the wrapper, and trying to blend them together and make it look 'good'. I imagine there will be a lot of metal impregnated epoxy involved in that. A boiler band is required near the front, and a ton of cosmetic bolts and washout plugs.

Then I'll get back to the reversing screw and see what I need to do to make it fit.

Then the boiler barrel cladding and cab roof to get this horrible sheet metalwork over and done with. Lots of details to hang off the sheet metal after that of course.

[ettingtonliam]

Looking at your work so far, I just can't imagine that anything you do, above or below the running boards, will be 'rubbish' Keep going, to use a favorite Irish expression 'Sure, it'll be grand'

[David]

Thanks, to be sure it will! As my father-in-law often tells me, the black paint will hide all the sins. We'll see. I removed that part of my post in any case, it was a pointless comment.

I forgot to say I'm pleased to relate another member of my club now has his version of this loco running on air. Another close by was looking really good until the owner bought a business, which stopped play for him.

[Deleted]

Apr 26, 2019 10:36:44 GMT ettingtonliam said:

Looking at your work so far, I just can't imagine that anything you do, above or below the running boards, will be 'rubbish' Keep going, to use a favorite Irish expression 'Sure, it'll be grand'

I agree with the above, it's going to be a very fine looking model David and I bet it will run just as good too...keep up the good work sir...

Pete

[Jim]

I would agree Pete the cladding is excellent and as the cladding is added so the loco starts to emerge and a very typical NSWGR B class is starting to emerge in your workshop. I think your cladding to date is up there with the best of them.

Great work David,

Jim

[Roger]

Apr 26, 2019 9:55:08 GMT David said:

Thanks Roger, I may try them. Do you think they might have bits chip off them? I told a few club members and one of them was wondering if the hardness could be a problem.



I very much doubt it, they are used in highly stressed ball races. That application is brutal to the surface so I can't imagine anything we might do to it would cause it to chip.

[David]

Thanks for the kind comments.

The photos must be flattering. The 'throatplate' is marginal and the firebox wrapper is looking better than I expected but the bends halfway down the sides are more like creases. There are no reverse bends below that either, I've not managed to make that work yet.

It has turned out better than I expected so far other than the throatplate, but my expectations were rock bottom.

Onto the transfer punches this morning if I can't get 2nd son to come and do his paper run. I also have to check the shower taps, my wife insists they are leaking. They have ceramic valves in them which are meant to last forever and not leak - anyone know about them?

I'm sure a pair of shiny safety valves will be a nice distraction on top.

I don't have rollers so my plan for the barrel wrapper is use a 'rolling pin' when it is on something soft. It will be 0.7mm steel like the firebox wrapper, which bent very easily. Do you think this will work?

[Jim]

Hi David,

I would think someone in the club would have a set of rolls you could pop round to use. The method you describe has been used in the smaller scales and could work with .7mm steel though I'd go with rolls if you can.

Jim.