NSW 'B' class mogul

[JonL]

I've had a few days like that recently. And then sometimes, the planets align, and you remember why you enjoy it so much.

[don9f]

Hi, if I was cutting a 2 start LH thread and starting the cuts near the chuck and cutting to the right and assuming the top slide is in its normal position, I would move the top slide handle anti-clockwise during the setup for the first thread, then once complete, move in the same direction to setup the second thread. I usually add a small amount of top slide for each pass to keep the cutting force biased towards the leading edge of the tool. In the case of LH threads, this would be the right hand edge of the tool.

For 2 starts you would have to do the same for each thread, so the top slide remains (in your case 0.75mm) in the correct relative position for each thread at the end of cutting.

Hope that makes sense

Good luck with your next attempt

Don

[David]

Another tool was made last night, this time out of silver steel and holding the blank in the toolholder for better support. Silver steel allowed me to file relief on the flanks of the tool before hardening. I tried to make it ever so slightly undersize although measuring this accurately is impossible with something this small and cheap ALDI calipers.

The helix angle was calculated as 6.8deg and a digital protractor (one of those boxes with a magnetic base) employed to approximate this.

I picked up the thread from the aborted trial, finished the first start and did the second and it looked better than Sunday's efforts so I had a go at the full length.

The wheels (literally) came off at some point when a change wheel unscrewed itself and fell off! I imagine all the running in reverse to get back to the start of the cut loosens the screws. I can't use the thread dial because it's an imperial lead screw and I'm cutting a metric pitch.

Anyway if I can make a tap and nut that will run on this screw I think it is good enough. It looks too fine to me but I counted the thread crests on mugbuilder's photo of his and I've only ended up with about 3 more I think.

It took 3 hours from the start of making the tool to the final cut on the screw - 18:00 - 21:00. That's blazingly fast for me.

I think the thread crests look better on this one than the earlier ones, both in more consistent width and a better edge. Still not great but hopefully workable. My worry is that the crests on the nut will have to be very thin and weak.

[Jim]

Looks the bee's knees to me David.

Brilliant job.

Jim

[JonL]

Had I seen that without context I would have assumed it was a commercially made bit of kit. Very nice.

[terrier060]

Feb 12, 2019 0:46:24 GMT David said:

Another tool was made last night, this time out of silver steel and holding the blank in the toolholder for better support. Silver steel allowed me to file relief on the flanks of the tool before hardening. I tried to make it ever so slightly undersize although measuring this accurately is impossible with something this small and cheap ALDI calipers.

The helix angle was calculated as 6.8deg and a digital protractor (one of those boxes with a magnetic base) employed to approximate this.

I picked up the thread from the aborted trial, finished the first start and did the second and it looked better than Sunday's efforts so I had a go at the full length.

The wheels (literally) came off at some point when a change wheel unscrewed itself and fell off! I imagine all the running in reverse to get back to the start of the cut loosens the screws. I can't use the thread dial because it's an imperial lead screw and I'm cutting a metric pitch.

Anyway if I can make a tap and nut that will run on this screw I think it is good enough. It looks too fine to me but I counted the thread crests on mugbuilder's photo of his and I've only ended up with about 3 more I think.

It took 3 hours from the start of making the tool to the final cut on the screw - 18:00 - 21:00. That's blazingly fast for me.

I think the thread crests look better on this one than the earlier ones, both in more consistent width and a better edge. Still not great but hopefully workable. My worry is that the crests on the nut will have to be very thin and weak.

Nice work. I think I put more of an angle in the sides of my thread.  Yours looks a bit vertical if you understand what I mean, but that maybe the angle the photo is taken.  I think it is roughly 14 degrees and would make the thread easier to match with the female nut.  Fun to machine though aren't they, especially a two-start!
Ed

[gwr14xx]

David,
If you have an imperial machine, wouldn't it be easier to cut an imperial thread - 1/8" pitch (8tpi) instead of 3mm - then you can use the thread dial indicator to re-engage the leadscrew after each cut. Also, you will probably get a better finish on a square or Acme thread if you dont have a 'full form' tool - say .5mm wide instead of the full .75mm. Set your compound slide parallel with the bed, then use it to move the tool sideways 1 thou at a time until you generate the correct thread form. With the compound slide set parallel, you can also use it to set the tool over by half the lead to cut the 2nd start. By the way - your estimate of 6.8degrees for the helix angle may be a little low - remember the helix angle increases as you get down closer to the root of the thread.
Another observation - the thread you have cut looks to be a bit on the thick side. Bear in mind that the nut that runs on it is going to be much shorter, and therefore more prone to wear, so making the male thread somewhat slimmer would leave more 'meat' in the nut - making it more robust and longer lasting.

Regards,
Eddie.

[David]

Screw, tap, nut! I was surprised but extremely pleased and relieved when the nut went onto the screw.



I'll describe making the tap in the thread in the tools section. I did spent a bit too long turning the tap the wrong way at first, forgetting it was left-handed and wondering why it wouldn't even start despite the oversized hole.

I have no idea how long the nut will last as it is brass with a damned small thread depth but it runs along the screw beautifully so for now I'm very happy with it.

Thanks Eddie. Most of what you say is beyond my ability for measuring and accuracy. I did use the compound slide for moving between starts, it's the easiest way when cutting a metric thread on the imperial leadscrew. I see now I could have used an imperial pitch but the first definition of 'square thread' I found said metric so that's what I did.

Because my lathe can go in reverse I find it easier to leave the half-nuts engaged no matter the thread pitch or type.

[David]

Feb 13, 2019 9:26:28 GMT terrier060 said:

Nice work. I think I put more of an angle in the sides of my thread.  Yours looks a bit vertical if you understand what I mean, but that maybe the angle the photo is taken.  I think it is roughly 14 degrees and would make the thread easier to match with the female nut.  Fun to machine though aren't they, especially a two-start!
Ed

Thanks Ed. You are correct, the thread flanks are vertical because it is a square thread, not trapezoidal. It didn't seem worth trying to make a trapezoidal thread cutting tool without a T&C grinder. I'm sure others could do it but it looks too hard to me. My bench grinder is sitting on a toolbox, not even bolted down never mind with any sort of usable grinding table in front of it. That means I can't even dress the wheel square.

Cutting the male thread holds no great mystery for me now I know how to compensate for the compound slide backlash properly. The difficult part is making the cutter, not doing the cutting. I decided it was easier to make a tap than try to make an internal threading tool - it's just too small.

[suctionhose]

Feb 13, 2019 9:26:28 GMT terrier060 said:

Yours looks a bit vertical if you understand what I mean, but that maybe the angle the photo is taken.  I think it is roughly 14 degrees and would make the thread easier to match with the female nut.  Fun to machine though aren't they, especially a two-start!
Ed

A piece of trivia: Acme's are 14.5 degree angle (29 deg included) as are racks for older style spur gears having 14.5 deg pressure angle. Why?
Sin14.5 = 0.25 A ratio of 1 to 4 was easy to draw or layout on a gauge.

[Roger]

Feb 14, 2019 1:00:52 GMT David said:

Feb 13, 2019 9:26:28 GMT terrier060 said:

. It didn't seem worth trying to make a trapezoidal thread cutting tool without a T&C grinder.

I think you could 3D machine that on the CNC without much difficulty with the tool pointing upwards and machining it on the end with the stock tilted at an angle.

[terrier060]

I love the reason for the angle Ross - so typical!

[terrier060]

Feb 14, 2019 1:00:52 GMT David said:

Feb 13, 2019 9:26:28 GMT terrier060 said:

Nice work. I think I put more of an angle in the sides of my thread.  Yours looks a bit vertical if you understand what I mean, but that maybe the angle the photo is taken.  I think it is roughly 14 degrees and would make the thread easier to match with the female nut.  Fun to machine though aren't they, especially a two-start!
Ed

Thanks Ed. You are correct, the thread flanks are vertical because it is a square thread, not trapezoidal. It didn't seem worth trying to make a trapezoidal thread cutting tool without a T&C grinder. I'm sure others could do it but it looks too hard to me. My bench grinder is sitting on a toolbox, not even bolted down never mind with any sort of usable grinding table in front of it. That means I can't even dress the wheel square.

Cutting the male thread holds no great mystery for me now I know how to compensate for the compound slide backlash properly. The difficult part is making the cutter, not doing the cutting. I decided it was easier to make a tap than try to make an internal threading tool - it's just too small.

Absolutely right David - I did the same - made a tap and it did the job well as yours seems to have done.  I wonder how difficult the wheel is going to be to turn.  Mine is quite stiff, the two-start thread acting like a high gear on the reach rod and valve gear.

[David]

I made a split collet to hold the screw shaft in the v-groove of the vise so I could machine a square on it. Worked well.



And here is it mostly finished. I unsoldered the front part again because it was clearly skewed. I have no good silver solder left so I just cleaned up the solder that was on it, smothered everything in flux and remelted what was there. Now it's not sitting on the base but enough is enough, I'll fill the gap with JB weld or similar.



I will mill down the bottom of the nut and solder a new piece of brass on so I can make the guides reach down alongside their rail. It works as-is but I don't like how it looks.

I have since thought of many ways I could have built these parts better. As far as I can tell I did it in the most difficult manner possible, adding many hours of unnecessary machining and fitting. But it is done and seems to work so the only thing I can do to put off the firebox cladding any longer is to do the handwheel.

[David]

I spent a few hours yesterday flattening the bottom of the reverser nut, sizing and soldering a new bit on, and milling the rebate again to make the guides deeper. All this proved was that the screw is still badly out of alignment and the nut won't sit flat. I'm trying to convince myself to live with it because the alternative is to unsolder the assembly and redo it but given I've not managed it so far it seems a bit pointless unless I'm willing to start using alignment pins and screws etc.

I watched this video the other night and with my new 6mm carbide cutters I decided to try and cut something on the CNC.



So I sketched up the curved angle the goes on the sides of the cab because I can't find any K&S angle locally to make it from. I used Stefan's feed/speed numbers of 36mm/min at 1300rpm with about 4.5mm doc / 3mm width of cut because it was cutting parallel to the edge of the stock. The chips were blue, the MDF was smoking, and despite slowing things down cutter broke within about 50mm. I admit this was a carbide cutter that had been used before, but I was hoping for better.

So I stuck a brand new one in, changed the params to 25mm/min, 1000rpm, 2mm doc. This worked ok but is killing the bottom 2mm of the endmill. It took about an hour to cut the part. I didn't want to come in at full depth with some 3D clearing op because it would have milled 99% of the stock away. On the second one I did some finishing passes that were a waste of time.

The CAD drawing of the cab was done in Draftsight years ago and imported to F360 recently so I could work with it. The profile that came in from the DXF isn't a true curve but a series of short straight lines. Don't know why, I'm sure I didn't draw it that way. The first part I cut had these straight lines which I hoped would be unobtrusive but they are really obvious even after a fair amount of filing.

I went back into the drawing and make a spline curve that went through each point where the straight lines met and this gave the correct curve back. There was still a trick in the extrusion operation in that F360 still saw the lines and was using them as well a the spine curve to decide where the profiles were, and I couldn't get rid of the lines because of the way I'd made a sketch from the cab side body - they were not 'lines' I could select and delete. So I had to select all the profiles that were chopped up by the lines and then zoom in and select the ones outside the lines/inside the spline.

The second attempt worked out pretty well, and fits the cab side really well. It needs filing down but I didn't know quite how tall to make the angle and didn't want to risk making it too short. I'm pretty sure this looks better than what I would have ended up with trying to bend even thin brass angle.

The photos show the two attempts and how obvious the facets are on the first one.

You can see how ragged the cuts from the endmill are - this was brand new and has cut exactly two of this shape. So I'm still doing something wrong. This is intensely frustrating.

[Roger]

Great results David, the leadscrew is a triumph of determination.

With regard to the issues with the milling cutters, my advice is always the same. Slow down the feedrate and RPM and don't go in too deep with a cut. Without flood coolant there's a limit to how much you can lubricate the cut and get the heat away. If it's smoking and blue chips are coming off, you're pushing it way too hard in my opinion. Rolled stock is always going to be harder to machine than free cutting Mild Steel bar, so you have to be even more cautious.

I always rough and then finish everything, you can't expect to get a good finish when you're removing a lot of metal, and the tool will spring so getting the finished size in one pass is not good practice. For a 2mm cutter, I'd leave 0.1mm to clean up and move that up as the cutter size increases. I use about 0.2mm on 6mm cutters going up to 0.3mm on bigger ones.

I know it's tempting to use deep cuts to get more value out of the cutters, but in my experience it leads to more trouble. If you start with say 15-20% of the tool diameter as a starting point for the depth of cut, I think you'll have a lot more success.
Don't be afraid to really slow up the feedrate either. 2mm cutters are much happier if you go at 10mm/min than 30, especially if you don't go too deeply.

In the end, you'll get way more life out of the cutter if you just accept that it's going to take a long time to machine a part rather than try to bully it to get the job done quickly. If you take deep cuts at speed, you'll just end up breaking the cutter or making it unusable. With gentle shallow cuts the tools last much longer.

[David]

Thanks Roger. I was really pleased with the screw.

The 6mm cutter 'sounds' pretty good to me at 25mm/min 1000 rpm, and the chips are not coloured. The only noise I hear that I don't like I'm sure is when it recuts chips. That's a little 'pink, pink' noise. I was blowing the chips out occasionally but didn't stand there doing it constantly.

I have to do the replacement for the faceted one so will use exactly the same parameters and see how it goes.

I'd like to have seen how those cutters in Stefen's video were performing after 30 minutes, but they were clearly ok for the short amount of time he ran them.

While I was making the brass nut above I was taking 6mm deep, 2.5mm wide climb milling cuts with a (used) 6mm cutter on bass on the manual mill at I think 1600rpm and it was going through like it was butter. I was going quite a bit faster than 30mm/min too. It's much easier when you can feel what's going on.

[Roger]

Hi David,
How deep was the cut with the 6mm cutter? It makes a huge difference as to how happy it sounds. Don't worry about it recutting swarf, you can't stop it completely, even commercially. If you clean up before taking a fine finishing cut, there's very little swarf to be re-cut, and it's fine enough not to ruin the finish. The key is to always Rough then Finish in my opinion. That's what I believe they do commercially, and all of the CAM outputs are explicitly set up to let you do that. You can double or treble the depth of cut, increase the feed and RPM for the finishing cut since it's not putting much load on the cutter. That way the finishing cut takes say 15-20% max of the total cutting time. The benefit of doing this is not only do you get a better finish, but you can also control the final size.

If you're trying to make something to an accurate size, just leave a bit of stock on for a pre-finishing cut to take the spring and runout of the tool into account, measure and adjust the amount of stock to get the exact size. It's no different to what you'd do in the lathe.

[David]

At the moment I am using a 2mm depth of cut with the 6mm cutter. I was trying 4.5mm deep or so when the cutter snapped and that was on the edge of the stock, not a full width cut.

I did another one of these angles yesterday to replace the faceted one and was able to leave it unattended. Right on 50 minutes. That's seems too long for a part like this if I was doing it for money but ok for myself. I was tempted to go for a 3mm depth of cut but I was sure 2mm would work unless the cutter was blunt enough to break so I left it alone.

F360 leaves stock for finishing by default and it is simple to set up a finishing pass by duplicating the roughing pass, unchecking 'stock to leave' and 'use multiple depth passes', and tweaking the feed/speed if you like. I didn't feel the result I got from the finishing passes was worth the extra time they took in this case. This isn't free machining steel either, just some black flat bar I found in my heap.

[Roger]

Feb 18, 2019 22:12:10 GMT David said:

At the moment I am using a 2mm depth of cut with the 6mm cutter. I was trying 4.5mm deep or so when the cutter snapped and that was on the edge of the stock, not a full width cut.

I did another one of these angles yesterday to replace the faceted one and was able to leave it unattended. Right on 50 minutes. That's seems too long for a part like this if I was doing it for money but ok for myself. I was tempted to go for a 3mm depth of cut but I was sure 2mm would work unless the cutter was blunt enough to break so I left it alone.

F360 leaves stock for finishing by default and it is simple to set up a finishing pass by duplicating the roughing pass, unchecking 'stock to leave' and 'use multiple depth passes', and tweaking the feed/speed if you like. I didn't feel the result I got from the finishing passes was worth the extra time they took in this case. This isn't free machining steel either, just some black flat bar I found in my heap.

Personally, I'd never use 2mm deep cuts with a 6mm cutter when roughing unless it was in Aluminium or Plastic. You end up having to go much slower to cope with the extra depth, so you don't save any time compared to taking half the depth at twice the speed. Yes, you use more of the cutting edge, but it makes it more difficult to get the material out of the flutes. I sometimes take 3-4mm deep finishing cuts, but then it's only removing 0.2-0.3mm of material.

Alibre can set the cut width just like you can with Fusion360 but it's not intended as a finishing cut. It's not an efficient way to finish the part if you're using the same feeds, speeds and depth of cut as the roughing pass. Maybe you can change those within the one machining operation on Fusion360, I don't know.
You can't on Alibre, so I always use a different strategy for finishing so it completes quicker. In my opinion you also don't want a sloping entry for finishing, that tends to leave a witness on the work. A 2D entry gives a better finish at the entry and exit.
I don't think Fusion360 operations can be intended for Roughing and Finishing in one operation, simply because these are often performed by different cutters, say a Ripper for roughing and a conventional one for finishing.

I think you'll see a significant reduction in machining time if you adopt these different separate strategies for finishing.