NSW 'B' class mogul

[Roger]

Jan 3, 2019 11:20:29 GMT David said:

The two blunt ones were carbide, but that's all the carbide I had so I had to fall back to HSS. I'll buy some more carbide cutters soon, I'm trying to find some ALTiN on eBay Australia. There is an Australian company that makes them but the prices are not suitable for learning with!

I doubt it was free cutting stuff, it was something I bought as an offcut years ago. I'll price some 12L14 next week.

The feed was 50mm / min, spindle speed 1000rpm. Do you just accept cutting times of 3 or 4 hours for a clearing operation as normal? That's where I am at the moment when things go well!

I might just give the CNC a miss until I have the cutters and free cutting steel. There is the rear screw support to do which is a slightly simpler shape that might be a good project when that stuff arrives. It's only 5mm deep too, not 10 so will be quicker.

Hi David,
That would probably explain the finish, although I think 50mm/min is too fast. I'd start at 20mm/min and expect to go up to 30mm/min. I use that kind of feedrate with whatever depth of cut I can get away with. I'd start with 0.5mm cuts and go up to say 1.5mm if it's cutting nicely.

Yes, a few hours to clear something is entirely satisfactory in my opinion, it's not commercial machining. You can't push it if you've got a light machine and no flood coolant and expect to get smooth cutting, it's just not going to happen. Once it's happy though, you can start it off and forget about it. That's the way I work, I lock up the workshop and occasionally check in with the web cam to see it's not done something horrible or it's stopped.

Do check that you're not wasting loads of time cutting fresh air though. I'd approach rapidly to say 0.2mm above the work or only just clear of the side so the approach moves are short. If there are lots of cuts, I reduce the clearance height to 1mm so it's not moving more than it has to. I double the depth and feedrate at least on the finishing cuts so they crack on and get it finished.

The reason I'm relaxed about how long machining takes is because in reality, you can't 3D model and program tool paths quickly enough to keep the machine busy anyway. Occasionally there's a backlog of machining to do, but it doesn't take long before it's all done and there's nothing new ready to put on the machine.

[suctionhose]

I've walked round the block twice trying to supress what I want to say but I still want to say it!

If couldn't make this part in under and hour (30 mins, probably) I give it up!

Method 1. turn it. part off, cut to shape, ss oil boss on.
Method 2. ss boss to 3mm sheet. Buzz it to shape on linisher (mill if you must)

CNC is a tool, yes, but it's not the only tool. Roger has demonstrated a mastery of the tool and that is one of his 'motivations' for building his Speedy.

There's an echo in my head: something about ..."wasting your time with files and hacksaws..."!!!!? If there was ever a tool devised for wasting time it is a computer!

Can we ever consider "appropriate tool selection" or "appropriate application of technology" or "return on investment of time" The trouble with believing computers are the answer to everything is too many eggs in one basket.

Building engines gives you the opportunity to learn / practice a plethora of skills that will let you make anything in an efficient way. Why spend hours and hours???

Sorry David, for choosing your thread to vent. Doing simple things in a complicated way makes no sense to me.......

[David]

No worries Ross, I sort of agree with you which is why I considered using the CNC machine for that little cab top strip 'over the top'.

I'm beginning to think CNC for many, but certainly not all, model engineering jobs is a waste of time unless you consider the drawing and programming, and the time it takes to become proficient is a part of your hobby rather than just a means to an end. I'm hovering on the border at present because everything takes so long with it and most things aren't going as well as I'd like yet and button pushing on the computer doesn't feel like material progress, but I see other people getting excellent results so it is just a case of putting the time in if I want it to start working. I have even had some good results. Imagine the wheels you could make with it, or the patterns if you still wanted to cast them. I made the coupling rods on the B class myself but I wouldn't blink at making them on the CNC if there is a next time around and I knew how to do it. I love watching it drill holes, for what it's worth ;)

Anything flat is silly when you can lasercut it more quickly and cost effectively, unless you have a CNC machine sitting around doing nothing. Then, why not use it if you can figure out how to hold the stock down.

I chose to do this is particular part because it had a number of profiles and heights etc. You're right about how easy it is to make manually and I may still do that. I reckon it would take me at least 3 hours though - that's just the speed I work at. But equally, it looks like a good exercise to learn how to use the CNC - it's a tricky set of toolpaths to get efficient and I'm still wondering about how I can improve the current program.

It will make more sense if I can stop hovering over it too. I managed to make two other parts while it was faffing around yesterday so if the part it made had not required so much attention, and was acceptable, it would have been a win.

I am learning a lot of new skills building this loco. CNC programming is one of them, and one of the harder ones! The drawing is easy but making a toolpath do what you want and efficient is quite difficult. One day I'll have a spreadsheet of what parameters work for each material.

[David]

The base and central web/nut guide for the reversing screw stand were done today. The I-beam is over long at the moment waiting for the front screw support to be done so I can see how long to make it.

[suctionhose]

Agree with what you say David. In fact I was hoping some discussion of this nature would come out on the Motivation Thread. ie The reasoning that underpins the decisions.

Certainly if CNC programming, 3D modelling etc is part of your objective then so be it. The commentary becomes, "I wanted to do this on the CNC because I was experimenting with work holding or blah blah. The result was, XYZ so next time..."

I have absolutely no issue with that choice. Each of us do what we do to tickle some fancy or other.

For myself, be it OCD, Control Freak, Analism, Time Obsessed or just Experienced, the objective has to be clearly defined in my mind. "At the end, I will have X". The learning on the way, the allocation of time to different aspects of the project has to be weighed against achieving that outcome. If it doesn't help, don't do it! Easy to get distracted and not achieve the objective.

I have used 3D Modelling, 3D Printing services, laser services but only if it helps. For much of it by the time you drive around, people muck you about, the errors etc, it only slows me down. My boss said to me, as an apprentice, "If you're not cutting metal you're not doing anything."

Appologies once again for invading your thread. I was FAR MORE productive in every aspect of life before computers started helping!

[JonL]

Don't we all default to the route we are happiest with? I've got a parting off tool, but to my shame I still put a bit of wood on the lathe bed in case of slippage and reach for the hacksaw, especially with smaller diameter stuff. More than one way to skin a cat.

[David]

I bought the machine with the idea of making parts for sale, perhaps as a gateway to a career change. I can't see it happening from where I am at the moment. Right now I'd rather have put that money towards the mortgage and I don't think I'd miss the machine. But I've spent the money so I have to use it and something will come of it. If I don't learn to use it I'll look like a right idiot.

I think 3D CAD is a brilliant tool just because it allows you to confirm things should fit together. It doesn't feel much like model making and I only use it help me visualise what I'm making like the cab and reversing screw or because I need to program the machine, not for 'fun'.

Nobby, I don't know if this would work for you but I use the upside down parting tool from Eccentric Engineering. It's really good if your chuck isn't a screw on type and your lathe can go in reverse. I've also starting using MGMN200 inserts and they work well for me too. I'm not shy with cutting oil. And a hacksaw and bit of wood if I can't get rid of the last little bit!

As for computers, I think places like this are very important for motivation and learning. Even with such an active local club I like coming on here and having a moan when things go south, and seeing what others are doing. At least here if people don't want to hear the me complaining they can just not read the posts! I spend as much time in the workshop as I want already, I don't see this as a distraction.

[David]

suctionhose is going to love this! In my defense it is a learning exercise and I did learn a few things.

Member atgordon has been giving me a tutorial on 3d adaptive clearing toolpaths and they seem closer to what I'm looking for - full depth of cut, a small stepover, and they take note of where the stock is and don't plough straight in. Unless you forget to reset your work coordinate system after faffing about in the CAM software, when they do plough straight in and this happens:



Fix the program, put in the other 5.5mm endmill and after about 40 minutes with no intervention I ended up with this:



Pretty happy with that. I'd cut 0.5mm below the 'bottom' of the part to ensure I had lots of room to clean up machining marks. That was 5.5mm depth of cut, 0.5mm width of cut, 1000rpm, 30mm / minute feed. I was pretty cavalier with the settings for drilling the 9.5mm hole so the finish in there is shocking but a bush goes through it so no worries.

Over to the lathe to part off and that big hole is causing an interrupted cut and this happens:



So that box of MGMN200 inserts I bought recently isn't going to get much use. Following nobbysideways' lead I finish the job with a hacksaw. I notice the chuck moving - the interrupted cut had caused the camlocks to come loose!

Flushed with that success I redo the program for the other reversing screw support and it estimates a bit more than an hour cutting time. That's a lot better than a couple of days ago.

Off it goes with the same cutting parameters except the depth of cut is now 10.5mm (so twice as much) and it's making a horrible squealing sound. I back off the spindle speed and that doesn't help, nor does reducing the feed. I squirt some WD40 on and that stops the squealing, but this allows me to hear a chuck, chunk, chunk sound. I'm guessing that is the cutter bouncing off the job because I've now given it twice the workload or something.

Never mind, that cutter is probably already done for so I let it go, and keep squirting WD40 and blowing the chips off when it starts squealing. Meanwhile I'm cutting the excess off the back of first bit I made on the manual mill.

58 minutes later the 5.5mm endmill is given a reprieve and a 2.5mm one is used to clean up the fillets under the boss and the hole drilled.



It looks a lot better than the first attempt. Perhaps next time I'll try a 0.25mm width of cut at this sort of depth - assuming I don't have to baby it along it doesn't really matter that it will take twice as long. I'm back at work tomorrow so it has all day to do things.

There was one problem with the toolpath in that when it had done everything required, it then went and did a bunch of detrimental cuts at decreasing depths on parts of the boss. These achieved nothing except leaving bad cutting marks and wasting time. I saw them in the toolpath but couldn't get rid of them in CAM. I could have edited the g-code but I would have found it difficult to isolate those bits. Luckily I'd left 0.1mm for a finishing profile cut which cleaned these marks away.

The other thing I don't like about the toolpath is it faffs about doing curved cuts when a lot of the time I'd rather it just went straight. F360 talks about consistent cutting loads etc which I don't understand but it seems more complicated than it needs to be. If I was writing the program by hand it would be a lot different. I tried to find a toolpath that would follow just one of the lines and go in but couldn't isolate a line segment. Perhaps a separate line on the drawing would work.

I decided not to try and part this off tonight. I'll have to go back to my 'upside down parting tool' and go a good bit easier than I did last time. Not sure how I'm going to get the excess material off the back of this bit either because I don't see putting it in a milling vise to cut across the back will do it any favours.

I didn't think I'd get either of these parts made this weekend so not a bad outcome.



Thanks Tony (atgordon) for the advice!

Lessons learned:

1. Cut deeper than the part's depth to give room for cleaning up after parting off.

2. Don't drill the holes all the way through or the parting tool will break.

3. Doubling the depth of cut means I should reduce the width of cut?

4. If using part-on-a-stick make the cylindrical stock as small a diameter as possible.

[suctionhose]

You still making that part? Yawn...!

[Jim]

I'll stick with fabricating.

Jim

[Roger]

Hi David,
A mixed bag then really. Personally, I'd never take a 5mm rouging cut, let alone a 10mm one. I would probably not exceed 1/2 of the cutter diameter as a depth in any situation, simply because the deeper the cut, the slower you need to go to avoid chatter. I doubt if your machine is man enough to take advantage of deep cuts unless you're finishing. I'd only define a cut width if going deeper than say 1.5 x the cutter depth. Almost everything you've seen me machine to date has been machined without defining a cut width. It takes more time, and that second cut is unnecessarily slow if you're only taking a small cut to make it a bit wider. Personally I think it's a waste of time unless you really need to do it.

The other issue is that you can't use the deep cut strategy if you're cutting the part from stock that prevents you entering from the side. These days I almost always use a sloping entry from the top, simply because it works for every situation. Ok, it doesn't use all of the flute length, but you don't when manual machining either! Yes, it's wasteful to machine small parts out of bigger stock than necessary, but it takes no longer if you're cutting from the top. Any stray parts outside of the cut get parted off. My finishing cuts plunge vertically into the existing cut, never further out than that.

The adaptive cutting with all those arcs are not necessary for most jobs, but it will save a lot of time when you're machining internal corners with very small cutters. The issue there is that you can clear the bulk of the material with a large cutter, but that leaves a lot of material in the corners. If the software is smart enough, it can create a path that reduces the tool load as it gets into those corners. Alibre isn't that smart, so I often program a separate roughing program for the corners.

I'm surprised you managed to break the parting tool, were you using flood coolant? I've never had any issues with parting off under power regardless of how intermittent the cut is, but pumped neat cutting oil is probably partly the reason for that. I also wonder how fast your RPM was? If you have a Variable speed drive, that helps too by reducing the torque at low speeds. Using power feed means you get a smoother progression for each rotation, something that you can't mimic by hand, and that's not going to help with intermittent cuts. My 'parting tool' is actually a Grooving/parting tool, so it's more rigid than most parting tools. That means I can leave less to clean up afterwards, often meaning that draw filing is all that's required without further machining. I had to machine the bottom off the grooving tool holder because it was too tall to fit in my tool post.

It's a pity you're not close by, I'd show you what these CNC and lathe jobs look like on my kit, and you'd probably see straight away the difference in the feeds and speeds you're using. You'll home in on what works for you in time, but it's frustrating getting to that point.

[jon38r80]

If the chuck was ratling about Id suspect that more than the interupted cut. Those parting tools dont like a twist at all ( Broke my first one too, the new one is fine)

[Roger]

Jan 6, 2019 10:23:44 GMT suctionhose said:

You still making that part? Yawn...!

Hopefully, this will kill once and for all the idea that CNC requires no skill or human input to get a good result. Skills take time to acquire, but the effort will be worth it in the end when it's accomplished quickly and accurately without drama. You can't compare someone's 40 years of experience of manual machining with a relative beginners efforts using sophisticated modern equipment.

[JonL]

I see the value and appeal of many different techniques to achieve the same task; I see some people making one component from two simpler ones soldered together, others machine CNC, others with a manual machine, and in the case of my Stuart 10V I did quite a lot through filing alone as I wasn't confident enough to use the vertical slide back then!

Diversity in approach is one of the key reasons I read this forum; I don't use every technique I read about but cherry pick those which interest me! Vive la différence!

[suctionhose]

Jan 6, 2019 20:09:50 GMT Roger said:

Hopefully, this will kill once and for all the idea that CNC requires no skill or human input to get a good result. Skills take time to acquire, but the effort will be worth it in the end when it's accomplished quickly and accurately without drama. You can't compare someone's 40 years of experience of manual machining with a relative beginners efforts using sophisticated modern equipment.

Touché Roger!

[David]

Jan 6, 2019 12:03:00 GMT Jim said:

I'll stick with fabricating. ;)

It would certainly be quicker for most people. But I've spent the money so I have to use it!

[Jim]

Can't argue with that David. I do the same.

Jim

[David]

Despite taking the scenic route I got there in the end. The front bracket isn't sitting on the base because I left the web below the boss long so I had the option of cutting a slot in the I-beam to help locate the bracket while silver soldering.

[Roger]

Excellent, a great result!

[David]

I've been working on the cab.

First up I found a number of holes on the drawings I'd missed when I converted them to CAD for laser cutting. So I drilled them.

The design calls for the spectacle plate to be silver soldered to the sides. I decided this would probably go badly so have spent time making angles to hold them together. It's taking days longer but I'm less likely to ruin the whole thing. The angles are flush riveted to the cab sides. Some of these are not quite flush and will need filling. There are a bunch of 10BA bolts in the spectacle plates that were dummies but now screw into the angles. Beneath these bolts, but on the same vertical line are some 3/64 rivets, 3 of the 5 will be riveted to the angles. I want to get some 1.5mm perspex for the glazing before I commit to putting the cab front and sides together permanently.

I finally cut out the opening for the reversing screw. I'd been putting this off because there are not many ways of getting it wrong that are recoverable. So far it looks good.

The commercial angle isn't 90deg so the cab sides are not perpendicular to the spectacle plate. I'll sort that out when I add the tabs to screw it to the running boards. Bending them out a bit isn't helping the flush rivets stay hidden.

I also finally drilled a hole in the plate that holds the two halves of the spectacle plate together to match up with the centre hole up near the roof. I also made the 6 clearance holes associated with this fixing slightly larger to allow the two sides of the spectacle plate to be better aligned.





The bracket for the reversing screw has been a bit of a trial. I cut some angle to form the L shaped front and side and some 2mm steel for the top. I milled the angle square and drilled the mounting holes - prematurely it turned out - and tried to silver solder the top on. Four times. The first two times the flux didn't look like it was working properly and the solder wouldn't flow no matter how much heat there was. I found some different flux, the tenacity stuff I used on the boiler and it seemed to be cleaning the steel but the solder still wouldn't flow or adhere properly.

The bracket was in pretty bad shape by now so I decided to play with some scrap steel. That was when I noticed the packet the solder was in said "Pro 15" - 15% silver. I tipped it all out and all the rods were 2mm, except for a couple of small scraps that were 1.6mm. I decided these were probably 55% stuff and tried it on the scrap and it worked like it should.

So then the bracket was soldered together with no fuss. A couple of hours in weak phosphoric acid cleaned it up pretty well. That makes all the scale and whatever flux hasn't fallen off just wipe off under running water. But you risk eating away at the part if you forget about it!

It was looking pretty poor at this point, not at all square any more. So I cleaned it up with a ripper mill and a light sanding. You can't see it because it's in the front corner of the cab. Note the inside of the bracket hasn't been touched except a bit of sandpaper in the corners. The acid is pretty good.





At this point it looked like there was a gap of about 1mm to the front of the cab. I was bracing myself to make a new one when I noticed the cab front wasn't really hard up against the angle that holds the cab together and that seems to have been most of the problem. I probably took 0.3 - 0.4mm off when cleaning up but that should be ok.

I'm waiting for the time I remember to make a fabricated bracket oversize so I can mill it back. Hasn't happened yet!