NSW 'B' class mogul

[Jim]

I agree with Roger, the loco is coming along very nicely indeed David. As to unwanted holes I've found automotive body putty does the job for me. It can be sanded to w nice smooth surface and it is heat resistant, we won't go into the details as to how I discovered that. I also use auto blade putty for minor scratches or fairing off features/shapes all hidden under primer, spray filler and  a couple of top coats. Now what little reputation I had will be shot to bits if the mugbuilder hears of this. 

Jim.

[David]

Yesterday I soldered the washout plug castings into the firebox wrapper. As a warmup/test I put some solder on the rivet I used to fill the bonus hole in the hope it would seep around the rivet and help blend it in. I used the smallest flame I could get on my propane burner and heating from the outside with the flux and solder on the inside it went very well. The steel has discoloured where the gal was sanded away but apart from that it seemed like the castings would go fine and they did. No solder made it through to the outside of the rivet though, so back to plan A which is ignore it until there is some primer on there.

Today I made another rear barrel support ring that was 1mm less eccentric than the last one. When I took the firebox wrapper off yesterday I noticed the washout plug castings in the barrel were fouling the rear support ring. I expected this but thought I'd got away with it because I couldn't see the problem with the firebox cladding in the way. So I cut some relief in the new ring and tried it before putting the firebox cladding back on to ensure it cleared the castings.

Then I spent some time fettling the throatplate. I sanded back the solder on the front of it so it was reasonably flat, hit it with a hammer some more to try and make the whole thing more flat, and then elongated and/or enlarged the bolt holes in it so it sat more square with the firebox wrapper. I think it looks better now. The fit between the wrapper and throatplate is still pretty ropey, particularly at and near the corners, but I can't make it any better and hope solder and filler will fill in the gaps.

I think the whole rear of barrel/front of firebox area looks a lot better than it did a few days ago.

The new band at the front of the firebox still needs to be done but I decided to quit while I was ahead. I'll try the old one again first anyway but I'm pretty sure it will still be crooked.

[mugbuilder]

Jun 14, 2019 12:09:17 GMT Jim said:

I agree with Roger, the loco is coming along very nicely indeed David. As to unwanted holes I've found automotive body putty does the job for me. It can be sanded to w nice smooth surface and it is heat resistant, we won't go into the details as to how I discovered that. I also use auto blade putty for minor scratches or fairing off features/shapes all hidden under primer, spray filler and  a couple of top coats. Now what little reputation I had will be shot to bits if the mugbuilder hears of this. 

Jim.

Tut Tut Jim,

Be carefull about how much primer and filler you use. Sometimes the Primer will seperate under the top coat and it will come off leaving half the primer on the job and half on the paint flake. Little is much better than more.

Barry.

[Jim]

Guilty as charged yer honour.  

Jim

[mugbuilder]

Jun 17, 2019 8:32:45 GMT Jim said:

Guilty as charged yer honour.  

Jim

I have tried several different filling and painting methods over the years, With fillers I have used auto plastic body filler which has usually blistered or cracked after a while. I have found Selleys plastic steel better as is can be filed and sanded to shape and even drilled and tapped if done carefuly but still can bubble if in a hot place. However they are in my opinion  both inferior to common soft solder if applied carefully. For years I have fiddled about with a fine flame and a stick of solder and had it run every where, but recently had great success usind a common old copper soldering iron that all plumbers usue to have . It was one that my dad had and looked terrible with a badly cracked handle. Those old copper ones are as good as new if cleaned up well. I used it to fill a throatplate to wrapper assembly very much like the one that Dave shows on his B class Mogul. I was able to build up a nice fillet around the joint that was then filed and sanded to shape. I NEVER go near Bakers flux as it is so acid and will often cause rust if steel is used. I always use 965 low melting point silver solder flux as it is very good with soft solder on steel or brass and is no where near as corrosive. It is more expensive but there is a lot at stake if rust appears down the track and the job is finished. With regard to primers , most auto types are chalk based and if used must be applied very thinly as they are prone to seperate through the primer layer and leave half on the job and half on the paint flake. I always try for the best possible finish on the raw steel or brass component to be painted and then use a thin coat of a good etch primer [that has no build qualities built in] and finish with the required paint. Car companies aim for a MAXIMUM finished paint build of no more than .004" Not only for economy but also durability.  NEVER USE HOUSE PAINT OF ANY KIND! Every time I find a good combination of paint brands they seem to go out of business or drop the product. These days here in Australia I am using Rhinogrip etch primer in black or grey [Excellent used alone on small fiddly bits anound brakes etc] and always use a good quality  Automotive spraying enamel for the top coat with a small dash of urathane additive if available. Catterpillar tractors put out a very good semi gloss black. I have painted over 80 locomotives so have had a bit of experience. Barry. 

[Jim]

Interesting and valuable advice and comments Barry, thank you.

I've found the old triangular stick solder as used by plumbers in days gone, to be good as a filler where heat wasn't involved. It tended to be thicker and less prone to spread everywhere.

Jim

[David]

It's interesting watching Barry paint something - he makes it look effortless, like anyone who's mastered something. Then you try it yourself and wonder if the laws of physics are different at your house.

I use a copper bit as described by Barry to coat my throatplate and also run a smear of solder over the inside of the wrapper where they overlap. I have two of them I got from my father's shed. The small one I used I could heat with a gas torch. Lord knows what it would take to heat the large one! The small one has lost its wooden handle but the steel rod didn't get hot. I didn't even clean the thing, I really didn't expect it to work.

The copper bit held enough heat to easily get the solder everywhere I wanted it and the throatplate looks a lot better now it's had the solder 'filler' and a sanding.

No work on the B class this week, work's been flat out. I won't have any time off until next Wednesday. The usual carry on:

1. This needs to be done last week!
2. 50% of the way though doing it - "we can't use technology X anymore, it's not reliable enough, scrap all that and use Y!"
3. 90% of the way though - the other teams didn't get very far on this project so we've cancelled it.

But you have to do your best until step 3. I expect that will be in about a month.

I haven't found the 965 flux yet. If I have trouble down the track I guess I'll live with it or do it all again in brass.

[David]

I got most of the way through making the forked end of the reach rod today. The way it's meant to be done is bend up a couple of bits of 7.5mm x 1.6mm steel and solder them to an odd shaped bit that sits in the end where they join. If you knew what you were doing that would take about an hour I'd guess, allowing for getting stock down to size and shaping the odd bit. It would take me all day, and it would look like rubbish.

So I still took all day but did it on various machines and came up with what I think looks many times better than what I would have got doing it as above. The bulk of the work was done using a 6mm rougher at 1400rpm and 90mm/min feed, 8mm doc and 1mm woc. It showed no strain. The finish passes were awful, the chatter and squealing were unbelievable. I tried both speeding and slowing the feed, to no avail. The finsh on the outside walls was atrocious so I changed the endmill before doing the inside and the finish was excellent but the noise was the same! Some sandpaper will sort it out.

I need to square up the internal corners and generally tidy it up but I'm pretty pleased with it.



Here is the outside profile and a couple of engraved lines done with a 45deg chamfer mill to simulate the joins. After this I took the piece out and drilled the cross-hole that will become the holes that go over the lugs on the reverser nut. The drill wandered a bit, natch, and on the side that will show! I'll live with it as long as it doesn't affect operation.



Drill out most of the waste. I used this website to set the feed and speed of the 3/16 stub drill and it was great. No centre drill required because it doesn't matter if they're a bit off. But the new stub drill went in straight.



That lot was cleaned out using the F360 trace toolpath. I could not get it to clear that space out using pocketing or clearing. A strip was left down the centre that I cleared out by jogging around manually. Then I drilled some 3mm holes up the front end to help create the slot.



A 2mm endmill, again with the trace toolpath, cut the centre of the slot to a depth of 6mm. 2000rpm, 60mm/min, 0.25mm doc. Manual jogging went down each side in 0.2mm woc increments.



Onto the manual mill to clear off the support base.



And this was the worst mistake. I should have cleared that bigger area to 1mm less than full depth to leave a support structure for this operation. As it was I'd cleared it to 0.75mm deeper, because on some job in the past it would have been better to go deeper. But not this one! I often make this sort of mistake and cause trouble for myself later. As it is I just got away with it. I pushed my luck to breaking point and it did go 'ping' right at the end (I wasn't going to go all the way to the end, knowing it would be trouble, but I couldn't help myself). But it didn't break or fly out of the vice! I'm quite surprised neither of those things happened.

I tried to find some packing to put in there but didn't have anything near the right size, and the next size up was 25mm so I'd have had to cut it down by half! It would have to be a perfect fit too, because the waste material was holding the legs firmly at whatever distance that was and there was no give in it towards the bottom.

[Roger]

Hi David,
I'm please to see that you managed to successfully make this part, it's actually a lot more challenging than it looks because of those long forks.

Personally, I wouldn't bother drilling anything out, it doesn't really help and takes time you could use on machining it all out. I don't know Fusion360, but it sounds like you might not have used a pocketing command but a profiling one if it didn't clean up. I use a 50% overlap on all my pocketing. Using more than that can leave areas that don't clean up. Maybe that's what happened instead?

I think the feed and speed for the 6mm cutter was way too high. I wouldn't machine those vertical edges at more than 30mm/min for the roughing and you might have to drop down to 10mm/min and say 800RPM to get a clean cut without chatter. One man's 'slowing down' is not the same as another!

One last point... if you see chips coming off blue, you're probably going too fast. I know some people like to see that, but unless you're in a race, I don't see the point.

[David]

Thanks Roger.

I could not get F360 to clear out that area using any sort of toolpath - 2D/3D pocket or adaptive clearing. I find it difficult to make it work in open ended areas like that. What I wanted to use was use a full depth adaptive clearing toolpath, like what John Saunders once showed for efficient slot clearing. I spent at least half an hour trying different things before I decided it was easier to draw a couple of lines and tell it to cut along them going down 1mm at a time. If I was making 100 of them I'd scour YouTube for clues to get it right, but it was taking longer to get the toolpath than make the cuts. I find many F360 toolpaths so difficult to make work that tracing along sketch lines is one of my most used techniques.

I only see blue chips with these face mills. If I saw that type of thing coming off an endmill I'd be dialling it back for sure. Next time I use it I'll put it back a speed and see what happens. It's a dramatic drop - about 1/2 the speed.

[Roger]

Jul 11, 2019 12:13:10 GMT David said:

Thanks Roger.

I could not get F360 to clear out that area using any sort of toolpath - 2D/3D pocket or adaptive clearing. I find it difficult to make it work in open ended areas like that. What I wanted to use was use a full depth adaptive clearing toolpath, like what John Saunders once showed for efficient slot clearing. I spent at least half an hour trying different things before I decided it was easier to draw a couple of lines and tell it to cut along them going down 1mm at a time. If I was making 100 of them I'd scour YouTube for clues to get it right, but it was taking longer to get the toolpath than make the cuts. I find many F360 toolpaths so difficult to make work that tracing along sketch lines is one of my most used techniques.

I only see blue chips with these face mills. If I saw that type of thing coming off an endmill I'd be dialling it back for sure. Next time I use it I'll put it back a speed and see what happens. It's a dramatic drop - about 1/2 the speed.

Hi David,
Fusion360 is way more sophisticated than what I use. All of mine is done the way you describe with sketches, either open or closed. I would have thought that you could draw a closed sketch that extended past the end of the slot for Fusion to use as a pocket. That's how I would have to do it in Alibre.
Half speed or less is fine when you have a cutter that's more than twice the diameter of the other ones you're using!

[David]

I might have been able to get the paths to work if I'd modeled a part with the stock below it so there was a floor to that area, or knew how to do what you describe which I think are called 'stock contours'. I've not had much luck with them yet.

If I had left a 1mm solid bridge or floor between the forked part and included it in the model this might have done the trick too.

I'll have forgotten all this by the time I make a similar part!

[Roger]

Since you know how to draw a couple of lines and get a profiling operation to choose those, surely you can draw a rectangle instead and point a pocketing operation at that? Fusion360 is very smart, so it hides the way it works from you. When you machine a 2D pocket, it almost certainly creates a profile around the top edge just like that. You don't need your sketch to follow the lines of the 3D model, it can be anything you like. All that's necessary is that it is a closed shape. It won't form part of the 3D model, it's just a 2D sketch in space.

[David]

You're correct! I created a sketch at the bottom Z height and drew a rectangle enclosing the area a I want to clear. I select the rectangle rather than some model geometry and it almost does what I want. It comes up with the usual (to me) bugbear of wanting to start in the exact opposite place to where I want - ie it wants to helix down into the closed end and cut to the open end. I want it to cut towards the closed end because there's all that waste material to cut straight into rather than helixing or similar.

Plus I think it would be safer, as part of the legs being cut would always have support in front of them.

Setting an entry point at the open end, changing the entry type, swapping between climb/conventional milling, none of it makes a difference. This is another problem I have a lot.

So I tried again with a different rectangle that was smaller than the area to see if an arbitrary shape could be used. It wouldn't do that. Then I tried a rectangle that I drew from outside the clearing between the legs, up to the 'front' corner where they join but otherwise touching the edges of the model, and it wouldn't accept that either.

F360 is clearly way smarter than me. I can't drive it. But it drives me - nuts. I'll try again next part. It's too late at night to be fiddling with it now. But thanks, I did learn something.

[Roger]

I'm surprised it wouldn't let you create an arbitrary shape, it may be a bit too smart! In Alibre CAM, you have to draw the sketch at the top of the area you want it to start at. In other words it knows the height of the sketch. I would try again, drawing the sketch at the top level of the cut. In Alibre, I tell it how deep to go, although it can figure that out from the model. I don't trust it.

I don't know if there are any controls in Fusion 360 for where the cut starts, you can't change that in Alibre. It's not normally of any consequence if you use a 3 degree sloping entry along the tool path. That will happily enter what there is material so it doesn't where it starts.

The only entry issue I have on simple 2D profiles is if I use a 2D side entry. In Alibre it can't always find a solution to that type of entry, so it doesn't create one. I have to sometimes use a linear entry and experiment with the angle until I get an acceptabLe path.

I get the impression that Fusion 360 is so smart that it gives the user the expectation that you can just tick a few boxes and it's all done for you. I'd suggest that you probably need to do a lot more than that to get it to do what you want. I was disappointed at how dumb the CAM module for Alibre was, but at least you can mostly bully it into don't almost anything

[David]

I just tried my idea of giving the area between the legs a floor and it made generating toolpaths a lot easier, and would have made machining the waste stock off a lot safer.

Here is the model of the final part. It doesn't have the engraved lines on it because they're just tracings over sketch lines. This body is useless for toolpath generation - it just confuses the generator and I could not get a single useful toolpath using this body.



To get the outside shape I used this body. It gives a continuous line around the bottom of the outside to work with for clear adaptive clearing (roughing) and outline (finishing) toolpaths.





Then I used the sketch lines to get the engraved lines. I could not get it to keep the tool down in a single run for the straight and curved part - notice there is a little lift where the lines join. I also could not figure out how to join the lines or make it consider them as a single entity. Perhaps if I knew at the time I could have drawn the curve with a long straight first segment rather than starting at the end of the existing straight line. I assumed that would not be necessary



Here is the body I should have used from this point on. This toolpath was made by selecting the tool, selecting the floor, pressing OK. It is exactly what I wanted but could not generate using sketch rectangles or the final body shape.





I have a lot of trouble getting F360 to generate usable paths working with the final body shapes. I assume there are ways to do it that don't involve creating intermediate bodies. Unfortunately there are times when you can't help but work with the final shape - radial toolpaths being one. The tender axleboxes I did a while back took a lot of faffing around. But I guess for the price, and considering the alternatives, it's still extremely good value.

[Roger]

I'm not sure why you need to leave a floor to get the tool path, on Alibre you would just tell it what depth to go to.
Are you using a pocketing operation on that? If so, it must be quite smart since it's an open shape and it has to deal with the centre line of the cutter needing to reach the open end. If that's all it takes to trick it, then you could make the bottom say 1 micron thick and get the same result. You don't need the bottom like that if it's being machined on stock that's supporting the fork from below.

It looks like you're machining all of that in one pass. I don't know how deep the cut is, but I wouldn't want to go more than 1mm deep if it's Steel.

I'd be wanting to spend a long session experimenting to see what tool paths it generates in a variety of simple parts like that so I could figure out how it works.

[David]

This is 'adaptive clearing' which I think is like pocketing but with a toolpath that is meant to keep a constant load on the tool and is meant to allow deeper depth of cut by using shallower width of cut.

I did the outside like this at full depth - 8.25mm, with 1mm woc (F360 defaulted to a 2.4mm woc which was too ambitous for me). It was fine with the rougher. I did the inside in straight lines of 1mm doc, 6mm woc. Given how the outside went I would at least try this toolpath at full depth and see what happened.

The 0.5mm floor in this serves a 2nd purpose by supporting the legs while milling the waste stock off the bottom. Then I'd just cut it off with the bandsaw and clean up with a file. That would be much safer than how I did it.

[Roger]

Hi David,
I don't think it matters what the clearing strategy is, I'm pretty sure that Fusion360 defines the areas to be cleared or profiles to be followed in the same way. I think the key to understanding the tool paths it generates is knowing how it extracts the boundary of the cut from the 3D model. For 2D machining, I suspect it draws a temporary sketch around the top of the clearance area, extending it by a cutter radius where there are open ends. If you chose a very small diameter cutter, you would be able to see that. In other words, an open end is a nuisance because you can't just draw a line across the end and treat it like a pocket because you would leave two fillets.

Personally, I would never take an 8.25mm cut on anything, I think it's asking for trouble. That's not only because the forces are large, but because chatter and tool deflection are often going be be a problem. You also have to clear that material out of the flutes. On a really large cutter that's fine, but on a small one it can lead to trouble. You probably won't be able to remove the material more quickly in any case, because you'll have to go much slower. The only advantage I can see is that you use more of the cutter's flute length. Do measure the part at the top and the bottom though, because there is likely to be a significant taper.
I use a 50% overlap on all clearing operations, and that seems to work well.

In the end, you'll have to see what works best for you. Some materials are more compliant than others, but I can't imagine the strategy you're using working on more challenging materials such as Gauge Plate or Stainless Steel. I'm pleased that you don't take anyone's word for these things, experiment and personal experience are much more valuable than following what others do, and this is how we make progress.

[mugbuilder]

Looks much better than mime Dave. A bit like using a sledge hammer to crack a nut. By not using screws to hold the proposed assembly together you will have to spring the fork over the reverser nut prongs. could make it harder to assemble.