Member postings for Andrew Johnston

Many years ago I had a wartime Taylor Hobson engraver, bought on Ebay from a seller who turned out to be a few doors down the road! He'd used it to engrave front plates for bus ticket machines, for those that remember the days of buses with drivers and clippies. And the old mechanical ticket machines where one set dials more complex than an Enigma machine followed by a satisfying whirr as the handle was rotated and a ticket popped out. As I recall the engraver ran at 18000rpm.

I rather like the idea (DC31k) of using a ballnose cutter as it will give a better look. But the cutters are rather more expensive than the straight endmills. Plenty to think about, but first I need to finalise the design of the rest of the regulator bracket.

Andrew

No need to apologise. I rarely get angry, and on the few occasions I do you definitely don't want to be in the same room!

Andrew

That may well be true of inserts, it isn't true of sintered carbide endmills. The sharpness of the edge is determined by the size of the carbide particles. They're pretty small these days. When I opened my first high performance 10mm cutter bought from Cutwel I carelessly sliced my thumb open. The cutter made an incision any surgeon would have been proud of; didn't half make a mess - blood everywhere.

Andrew

Thanks for the replies, plenty to think about. To avoid confusion this is the current CAD model of the part:

For scale the width of the part is 3/8" and the lettering is 0.08" high. Material is low carbon steel.

The supplier I was looking at was Drill Service, about the same price as APT and a quarter the price of the YG cutters from Cutwel! Sadly Drill Service don't provide any detailed cutting parameters. I found something from Harvey Tool where, for a 0.015" cutter, they recommend a chip load of 0.00009" with a depth of cut of 1D and radial cut of 0.13D for finishing in steel. They seem quite agressive depths of cut. Out of interest I looked back at my CAM files for the governor bevel gears (1mm ballnose in bronze) and the smokebox nameplate (1mm endmill in engraving brass) and I ran both cutters at 24000rpm and 350mm/min.

I'd thought about engraving cutters, but assumed they wouldn't be good for raised lettering; is that a misconception? Special thanks to Adam for his generous offer, which I'll keep in mind as I decide which way to go. I have some engraving cutters from Arc but they have a sharp point unlike the Taylor Hobson ones.

I had rather assumed that I'd need to pocket out most of the metal with a larger cutter and use the small one to re-machine just the detail. But oddly the CAM system doesn't think it'll take long to do the whole thing with the small cutter. Likewise using a tapered cutter and V-carve was suprisingly quick.

As an aside I use a program called NCPlot to sanity check my G-code. Run times are normally quite close to the CAM program estimate, and a bit more accurate. Today I've been CNC milling and drilling the valve chest covers. The CAM program thought each one would take 40 minutes, which I finally got down to 30 minutes by tweaking parameters. NCPlot thought it'd take 14 minutes, which it did. Not sure why there's such a big difference, although I don't normally do a lot of CNC hole drilling. So may be the CAM has it's knickers in a twist over drilling?

I run flood coolant on the CNC mill so getting rid of the chips shouldn't be an issue.

In the picture I have of the regulator on a full size engine the definition of the lettering is fairly poor, so may be I shouldn't try too hard!

Andrew

That sort of response always dampens the spirits, as it usually means the speaker is probably going to do exactly what they claim not to be doing.

I have no idea about high frequency vibration, but the high speed spindle has made a good job of machining so far with small cutters without breaking any. While I have some high end professional accelerometers I'm dis-inclined to muck about making a charge amp. Been there, done that and it's non-trivial for large signal bandwidth,

I have a golden rule for buying cutters. If I buy two I don't break the first one. If I only buy one I break one, usually on a Friday evening.

Andrew

Correct, supercapacitors have come a long way since the 1980s when they could barely run a real time clock IC or memory backup at a few microamps. Last year I was involved in a project where the client wanted the single board computer (SBC) to shut down in an orderly fashion after mains failure. Given the boxes were up lamp posts they weren't keen on batteries. We ended up with two supercapacitors in series, which kept the SBC going for the minute or so that it took to shut down while drawing several hundred milliamps. Analog Devices do a rather nice control IC with two switchers, one for charging and one for supplying the load, while also looking after all the housekeeping.

Andrew

I'm in the process of re-designing the regulator mechanism for my traction engines, I need to reproduce a cover with a 0.4mm deep recess that contains raised lettering; the original would have been cast. It's simple to draw and in theory to use the CNC mill to create the recess and lettering.

However, the small scale of the lettering means I need to use a 0.5mm slotdrill. I've used 1mm endmills and slotdrills before for engraving and gear manufacture, but not as small as 0.5mm, nor in steel.

I'll be running the cutter at 24000rpm, but the question is what chipload would be advisable and what depth of cut and stepover should be used. I've got some values for the parameters in mind, but what does the panel think?

Andrew

Ron is correct, with one minor bit of pedantry, the limit is below FL100, not quite the same as 10000ft.

Andrew

Silver steel or gauge plate will be fine.

Andrew

Nipped out into the garden this afternoon to do a bit of digging between trying to sort out my mum's car with the garage and getting a quote to vulcanise tyres on my traction engine wheels. Not sure which was the most painful! I'd only been out a few minutes when a Spitfire flew over at less than a 1000ft on the way to Duxford. Certainly shifting, I hope he was below 250kts. Later model with pointy wingtips and the engine sounding like it had something loose, so a Griffon.

Andrew

I've got the exact same model, given to me by the electronics assembly company I use. I don't know anything about the company, but the drill was designed to be used for drilling PCBs.

Andrew

Errr, we'd worked that out. Despite what you might think we're not complete numpties. We're talking about a different sort of tapering, as raised by Ketan. I doubt you're problem is due to the borinmg bar, more likely to be the insert and technique.

A helpful sequence for use on forums is read, understand, post.

Andrew

Neither, it'll be enough voltage to drive the rated current through the resistance of the winding. However, 0Hz is not applicable to simple VFDs. They have a minimum frequency of operation as well as a maximum. To maintain control at 0Hz one needs to use vector control rather than V/f open loop control. As an example for electric vehicle drives full torque at zero speed is useful. It's the equivalent of holding on a hill by slipping the clutch.

The V/f characteristic is quite reasonable as controls go, but tends to go wrong at low frequencies where the voltage drop due to coil resistance becomes a significant proportion of the back emf. Some VFDs allow one to modify the V/f characteristic, modelling it as a series of lines of varying slope. Torque boost simply means increasing the winding currents more than rated current in a controlled manner. And preferably for a short period unless the motor has extra cooling. It can be useful to get through a short term overload.

The experiment involving not being able to stop the chuck at low speed doesn't really prove anything. Below base speed, in simple terms, the torque is constant, and probably more than one can hold, whatever the speed. However, given that power is torque time angular velocity the power available for cutting is proportionally reduced. That's not useful if a large item needs to be turned at slow speed. That is why people say (correctly) that VFDs have some limitations at low frequency. A mechanical speed reducer (belts or gears) increases the torque at low speeds over and above that at higher speeds, keeping the power constant.

Andrew

Small and large brushes for moving swarf when cutting. When cutting a pocket a bendy straw can be useful to blow the swarf out. Bet the **** government didn't think about machinists before they banned bendy straws!

Cleaning up after machining done with brushes and dustpan and brush. The floor gets swept/vacuum'd as needed.

Andrew

Don't know the answer, but I'd agree that one generally gets a better finish on the way out, for internal bores at least. I haven't tried it on external tapers.

Andrew

That's what I thought as well.

It would be interesting to know what grade of stainless is being used. Of the austenitic stainless steels 303 and 316 are easy to turn, whereas 304 is 'orrid stuff. In my experience 303/316 are fairly insensitive to cutting speed and depth of cut, although I'd regard the depths mentioned as rather shallow. The only golden rule is never let the cutting tool rub, as the steel will work harden, 316 especially so.

I don't understand why a flexible boring bar would cut a tapered hole. There might be some bell mouthing as the cut comes on. But once the cut is established, and the bar has deflected by whatever it wants, then the cutting process is stable and the diameter of the work won't change. The diameter might not be what you expect, but it won't be tapered. That's definitely my experience. It's counter-intuitive but the best way to kill chatter when boring is increase the depth of cut and/or the feedrate. The boring bar may deflect a little more, but if the cutting process is stable then the bar will not chatter.

Andrew

Three things:

1. The insert may be new, but where did it actually come from?

2. The screw holdng the insert looks odd; they're normally Torx rather than slotted, and the countersink is not standard so an ordinary screw won't work

3. I adjust the gib on the compound to be slightly loose if anything, so there is no hesitation when feeding

Andrew

Well, so it is, on mine at least:

The nut behind the fixed jaw is in fact two nuts locked together,, and is clearly where the reaction takes place in contrast to the weedy collar at the other end. Generally I'd have said that shaper vices tend to be brick built outhouse style compared to machine vices of comparable capacity. Not sure if it's odd but the vice pictured has a square thread rather than Acme.

Andrew

Been there, done that. Twice over the years.

Lesson learnt, if I get in a tizzy go and have a cup of tea rather than keep rushing to get the job done. The client will just have to wait! Mind you one needs to wind the handle quite a long way in the wrong direction to break the stem. And I do use the Haimer most times I set something up on the manual and CNC mills.

Andrew

Have you considered one of these:

Although the ball is 4mm diameter one simply moves the ball until the dial reads zero, so whether the mill is metric or imperial is unimportant.

Andrew