Member postings for Andrew Johnston

Chris,

I've never managed to get the needle exactly stationary either, slightly less than one division is the best I've achieved. Rather than measure the four cardinal points by trapping a fag paper and reading the DRO I suppose I should measure using a more accurate means.

The ball on my device is a very good mirror finish, so I suspect it's not the problem. My device seems to be spring loaded, or similiar, so I'm also discounting internal backlash. That leaves the surface finish of the reference cylinder. I think this may be the problem. My reasoning is as follows. A 1µm Ra finish is quite reasonable for a turned finish. However, Ra is an average value, I believe that Rz, which is a peak to peak measurement, is now preferred. There is no exact conversion between Ra and Rz, but rules of thumb seem to be that Rz is generally 10 to 20 times Ra. So for a 1µm Ra finish that gives an Rz of 10µm to 20µm, ie, 0.01mm to 0.02mm peak to peak. In imperial half to one thou. I'm seeing an error of about half a thou, and slightly less than one division of movement on the dial. While the divisions are arbitrary I'm assuming that one division is on the order of half a thou, or 0.01mm. The manual for my device claims an accuracy of ±0.003mm, which would be difficult to detect if each division spanned a bigger value.

Regards,

Andrew

That must have been when I got bored and flicked to the next channel to watch a bit of 'Morse'.

Andrew

They are cheap compared to those from Blake or Haimer, for instance. I bought a Haimer one last year. Initial experiments were a little disappointing, but now that I'm getting used to it I reckon I can consistently get centred to within half a thou. I suspect that the limitations are now the finish on the bore one is centring on, and the issue of moving the mill table by a few tenths here and there.

Regards,

Andrew

Damn good questions; I've been wondering the same - Andrew

Edited By Andrew Johnston on 04/06/2013 10:28:11

Niko,

Funnily enough I had a vague recollection, when watching the programme, that there was a hidden chess master involved, but I don't recall it being mentioned in the programme. May be it would have got in the way of the central thesis being pushed by the presenter.

Regards,

Andrew

Gary,

Dormer endmills are over-priced in my view. I use their drills and some taps, but not milling cutters. My local tool emporium stocks Garr and Guhring. Garr is fine for aluminium but a bit soft for steel, for which I use Guhring. I see that Cutwel do 12mm HSS cutters from YG (Korean) for about £15, including VAT. A good proportion of 'professional' cutters these days are centre cutting, 2,3 or 4 flute, as industry largely uses CNC mills, where the traditional end mill makes it difficult to start cutting in a pocket, even with a helical or ramp lead in. For new cutters I've standardised on 6mm and 10mm three flute carbide centre cutting end mills.

Trust me, Chris's suggestion does work.

Regards,

Andrew

Well the presenter is a professor of the history and philosophy of science. There was a half-hearted attempt to explain cams, but sadly not much more. What would have been really interesting is to know how the articulation of the swan was done. The writing and drawing automata were amazingly smooth, I wonder how many prototypes were made before they worked? The programme also glossed over how the chess playing automaton worked, was it working to a fixed algorithm, or could it make decisions, and how did it know what move its opponent had made?

While I would agree that the automata had stored programs I think the presenter over-egged the comparison with the modern computer, in that he seemed to miss the point that a computer can make decisions based on input data, and can therefore follow a different path every time.

Overall, the programme had just enough to keep me watching, but it could have been so much more interesting.

Regards,

Andrew

That was my initial reaction too, but I think JasonB is right. But it is only true for workpieces that are much narrower than the diameter of the flycutter. Imagine we have a narrow workpiece and the diameter that the flycutter swings approaches infinity. The flycutter path across the narrow workpiece approximates a straight line more and more as the diameter increases. In the limit, at infinite diameter the path is a straight line, so the surface generated can't be concave.

Gary: If the 'stringy' swarf only appeared on the last cut, near the edge of the work then I wouldn't worry about it. I think it's the result of cutting near the edge rather than a problem with feeds and speeds. I sometimes see something similiar using very different cutters, much higher speeds and feeds and a different mill. My first thought on the relatively poor finish would be too look at the end mill. What brand is the end mill? I'll try and sketch something out on the width of cut issue, as I see it, tomorrow evening.

Regards,

Andrew

Wot? My assumption was as follows; let's assume that the 1000W quoted for the X3 is at maximum speed, and that power is proportional to speed. So, with a top speed of 1750rpm, a speed of 500rpm gives 285W available. I suggested an order of magnitude increase should be possible, ie, to 0.1 cubic inches per minute. A 12mm cutter full width at 2mm DOC and 100mm/min feed is 0.146 cubic inches per minute. So both are comfortably within the power available, allowing for a few losses in the drive train. As a first approximation the results seem fairly similar to me?

Gary: I think that 2 thou per tooth is perfectly respectable, particularly in steel. I only mentioned it as a common error is to think that going slow to start with is being sensible. As you have already twigged all that happens is the tool rubs, and then cuts, and then rubs. So you end up with chatter, a blunt tool, and a lousy finish. Can I assume that for 1rpm on the handle you mean 1rps, as 1rpm seems awfully slow! In terms of width of cut I normally use less than D/3 or greater than 2D/3. I've never had problems using the full width for cutting. In theory D/2 is to be avoided as it hammers the teeth as they enter a cut.

Interestingly I never run coolant on carbide tooling, except on the CNC mill for steel and aluminium, to get rid of the swarf. Carbide is quite happy running red hot, so let it. It helps with the power and finish too. To some extent the power needed per unit of material removed goes down as the speeds go up, particularly with harder materials. Of course you still need to have the horses available, which most people, including me, don't have. If everything is cutting as it should then most of the heat leaves in the chips; just don't get them on your hands, or worse down the shirt!

Regards,

Andrew

Hi John,

Thanks for the info. If I've got my sums right that means the table on your mill will cover it's full movement in about 2.5 seconds, and the rotary table will do one turn in 3 seconds? Both of those are way faster than my CNC mill, so I'm really interested in how they are achieved. From memory my CNC mill rapids are about 25mm/s on the axes and 30°/s on the rotary table. I must admit that my axis rapids seem pretty fast when you're driving a tool towards the table, although it can get tedious waitng for the rotary table to get back to zero.

If I'm correct the KX1 has a 4mm pitch leadscrew, so if the motor is direct coupled it's doing about 1500rpm to get 100mm/s? Are you using stepper motors or a servo system?

Regards,

Andrew

GaryM,

While speeds and feeds do not need to be slavishly followed, they can give useful guidance for particular combinations of cutters and materials. Knowledge of cutting speeds in low carbon steel may have saved a few ruined cutters in the recent slitting saw thread on this forum.

As a guide 1hp will allow the removal of 1 cubic inch of low carbon steel per minute. With the figures you mentioned I calculate that you're removing about 0.01 cubic inch per minute. Your mill should be capable of at least an order of magnitude increase on that.

For HSS steel cutters in low carbon steel a rough rule of thumb is 100sfm cutting speed. For a 12mm cutter that gives about 800rpm. So you could go faster, but 500rpm is fine, it's not that critical.

I'll make an assumption that your end mill is four flute, so at 500rpm and 100mm/min you have a load per tooth of 0.05mm, ie, about 2 thou. If the power is available you could probably double the feedrate, but it is perfectly acceptable as is, and is large enough to avoid chatter.

Personally I find that 'dabbing' coolant on is a waste of time; either nothing or flood coolant for me. However, I would agree that soluble oil is the way to go, and is what I use on all my machines. Tapping oil is designed to be 'sticky' so that it adheres to the tap, and holds the chips. Which is opposite of what you want for milling.

I also have one rule for cutting tools:

Never buy cheap cutting tools - they're a waste of money

Oh, and if anybody says that as amateurs we don't need to worry about feeds 'n' speeds my response would be that the metal doesn't know it's being cutting by an amateur on a small machine, so why not at least be aware of the information available.

I hope that helps, or at least provides food for thought.

Regards,

Andrew

A useful starting point for HSS tooling in low carbon steels is a cutting speed of 100sfm. So for the aforementioned 3" cutter an ideal speed is 127rpm, but slower is no problem.

Fine pitch slitting saws are ok for shallow cuts like slotting screw heads, but useless for deeper cuts, as the gullets get jammed with swarf. For deeper cuts always use the coarse pitch saws. To echo Skarven I normally aim for a chip load of at least 2 thou per tooth with a slitting saw.

Regards,

Andrew

Many years ago I bought a cheap set of metric solid dies and taps. They soon got recycled; it looked like somebody had ground the 'taps' using an angle grinder. There is no point in buying cheap cutting tools; they're just a waste of money.

Regards,

Andrew

That's a rare one, otherwise known as 1BA.

Andrew

Blimey, 100mm/sec, that's impressive, 6 metres a minute, better than a Haas TM3.

Andrew

It's easier to use integer arithmetic - no calculator needed.

Andrew

A simple calculation may be illuminating. Let us suppose that we want to move, and track, the milling table at a moderate 300mm/min, or 5mm/sec. Let us assume that we have a 5mm pitch leadscrew, and that we accelerate to 300mm/min in one turn of the handwheel over 1 second. So, over 5mm and 1 second we've gone from zero velocity to 300mm/min, or 5mm/sec. Acceleration is the change in velocity divided by the time. So, using metres, we get 0.005m/s/s. But 'g' is about 9.81m/s/s. So our acceleration is about half a milli-g. Which according to Russell is less than we can resolve anyway.

Regards,

Andrew

Hmmmm, I think you're one 'government' out, I think you'll find it was actually the previous (Labour) goverment that dumped night classes for mature students. They also decided we didn't need people with higher degrees, so they went too, to a large extent.

Regards,

Andrew

John,

Thanks for the hints, I'll have to give them a try. With the Coventry dies do you use the normal cutting teeth, or set the die at an angle so that the teeth that are normally used to guide the die are used to cut instead?

I'm puzzled as to why thread mills should be so expensive, after all they're not much more than a glorified tap. It could be partly explained by carbide versus HSS, but carbide end mills aren't that expensive. May be it's much more difficult to form grind carbide than HSS?

Regards,

Andrew

Neither have I in a practical sense. However, I do have the CAD model created, CAM toolpath created and G-code generated and checked. For a practical experiment I've been put off by the cost of thread mills, expecially if it goes 'ping'. However, I have recently ordered some single point thread milling tools from Maritool in the US. Watch this space for the results!

Andrew