Member postings for Andrew Johnston

For me at least the links in the OP lead to a page where I would have to log in to view. So I haven't been able to see the finishes for myself. But I apologise for wasting everyones time by making silly suggestions. It won't happen again.

Andrew

Feedrates are way too slow, leading to rubbing rather than cutting. With the parameters given I'd be feeding five to ten times the rates stated. Plus one for an aluminium specific cutter. Aluminium is one of those materials when climb milling on the final cut can be advantageous in getting a good surface finish.

Andrew

No need to apologise; can't imagine we'll be duelling at dawn!

Andrew

Swords at the ready!

That's not what I said; I don't use it for drill grinding, not saying it was of no utility. It's my personal preference to buy rather than grind. By the time my jobbers drills need replacing they tend to be rather a mess. I use the Clarkson attachment for grinding the lead on taps, both commercial and home made:

Andrew

I've got a Clarkson tap and drill grinding attachment, but never use it for drills. For standard jobbers drills, I just buy new ones as needed. By the time they're blunt they're worn out anyway and not worth sharpening. For drills above 12mm I grind freehand. Life is way too short to waste grinding a drill that might cost a couple of quid. I doubt I spend more than £30 per year on standard jobbers drills.

Andrew

I'd say go for it. Apart from dabbling with a loco in my teens, and a part built hit 'n' miss engine, these are my first engine build:

I've made everything, apart from the steel boilers and rubber tyres. Also produced a lot scrap, nearly enough for a third engine! Like Jason said I started at the front end, but moved onto the gears qjuite quickly, because I like gears. At any given time I make what interests me, rather than doing things in a logical order.

If I can do it I am sure you can!

Personally I wouldn't buy a part built engine, as one is never sure of the workmanship and what changes may, or may not, have been made. I'd rather mess it up myself than buy someone else's mess up.

Andrew

It might be worth checking that the lathe still turns parallel. I don't think the gap piece is guaranteed to go back in the same place. On my lathe (Harrison M300) I had all sorts of problems with tapered turning until I realised that I hadn't put the gap piece back properly. To go back properly everything has to be spotless, and then cleaned again. The screws also have to be tightened in the correct sequence.

Andrew

I have a recollection that Moore and Wright used to state that their radius gauges were machined, and hence were better than the opposition who punched them.

Andrew

The only one of my industrial machine tools that has a clutch is the horizontal mill.

Ah, just remembered that the power guillotine has an electromagnetic clutch.

Andrew

Any genuine modeller/engineer would know that ARC is based in the UK. Nuf sed!

Andrew

That's because the 5A fuse is too small. The plug fuse is there to protect the upstream wiring, not the downstream electronics. A fuse is a thermal device and even under overload conditions can take tens to hundreds of milliseconds to blow. The semiconductor devices on the board will blow orders of magnitude faster than that if there is a fault. Using a smaller fuse doesn't achieve anything, apart from increasing the PITA factor.

Andrew

It's not particularly useful to describe the feedrate on a lathe in terms of distance per unit time. Conversely it's not particularly useful to describe the feedrate on a mill in terms of distance per revolution.

I expect that the mill was originally designed for use with an overhead flat belt driving the spindle. Creating the feeds from the spindle would be simple if not ideal. Looking at the diagram above there should be four feeds which will vary in proportion with each of the main spindle speeds.

Andrew

I'm puzzled as to why the feedrate varies with spindle speed, it's not logical. My mill has a separate 1hp feed motor and gearbox, but on a mill with a single motor for both functions I would expect the feed drive to come directly off the motor shaft, not the arbor. Sounds like a fudge by a previous owner.

I'd be wary about using a VFD. Of course it is useful in order to run a 3-phase motor, but in this application I'm not convinced that the variable speed ability is useful. As the motor speed varies so will the feedrate, and as the motor speed is reduced below base speed (normally at 50Hz) the available power will also decrease. in proportion.

Andrew

Home made flycutter with single point HSS toolbit, 8.6" diameter, running at 54rpm in cast iron:

Roughing cuts were 50 thou deep and 15 thou per rev feed. For the finishing cut I slowed the feed to 4 thou per rev. Took ages as 0.004" per rev at 54rpm is 0.216"/min. Made worse as I was feeding in Y which means it was by hand - power feed on X only. Made slightly easier as the mill is metric, and since 0.004" is about 0.1mm it was one big division on the dial per clonk.

Andrew

Out of idle curiosity I googled "old dog", and it said you can't teach them new tricks.

Andrew

Correct. I was simply saying that the maximum feedrate on my mill is 16"/min, so a feed of 12"/min is at the high end of the feedrate range on my mill. For completeness the minimum feedrate on my mill is 0.4"/min.

A spindle speed of 250rpm seems awfully high for a 3" diameter HSS cutter. I'd be running at around 100rpm in low carbon steel, and lower still in cast iron or alloy steels.

As stated, the sales literature slowest speed of 66rpm is more like what I would expect for a horizontal mill. For comparisonj the speed range on my mill is 30rpm to 1200rpm, although it does have a dual speed motor.

Andrew

Those speeds are quite high for HSS side and face cutters. Let's assume a 3" diameter cutter. A rule of thumb for a HSS cutter in low carbon steel is 100 feet per minute. For a 3" diameter cutter that equates to 106rpm.

Feed rate = chip load x number of teeth x rpm

So for the numbers given, 0.004" chip load, 12 teeth and 250rpm we get a feedrate of 12". That's quite high, my (metric) horizontal only goes to 400mm/min.

Ideally the arbor speeds need to be reduced. I run side and face cutters, and slitting saws, from about 60rpm to 120rpm in steel.

Andrew

All cutters have some run out. On a horizontal mill the depth of cut is normally far greater than runout so it is not a problem. When I first used my horizontal mill it chattered, and the whole mill shook. Impressive given that it weighs the best part of 2 tons. The cure was simple; double the feedrate so that the cutter teeth were cutting, not rubbing. I use an absolute minimum chip load of 4 thou/tooth.

Andrew

Part 2:

As mentioned smaller, and possibly cheaper, VFDs often use V/f control. Below base speed the applied voltage needs to be reduced in proportion to the frequency to keep phase currents constant. Open loop control V/f works reasonably well, except at low frequencies. Here the voltage drop across the rotor resistance becomes significant, meaning that the applied voltage is too low, also reducing phase currents and torque.

As MikeP mentioned these issues can be solved with vector control. There are many variants. but essentially they all transform a complex rotating vector into a quasi-stationary 2D vector via the Park-Clarke transforms. This 2D vector has two components, d (direct) which is the flux linkage component of the phase current and q (quadrature) which is the torque component of the phase current. These can then be controlled with simple PI loops. Naturally this takes some computing power, but with the availablitily of cheap DSPs is no longer difficult. These algorithms do not directly measure the position of the rotor. Sensorless vector control can estimate rotor position, provided the rotor is turning. Control can typically be maintained to a lower frequency than a V/f controller, down to about 1Hz. These algorithms can maintain constant phase current, and torque, down to low frequencies. So in that sense they are better than V/f. What they cannot do is magically increase torque at low speed. Of course they can provide a short term increase in torque, but that is not sustainable long term due to motor thermal issues.

For full torque at zero speed vector control generally needs an external measure of rotor position. Full torque at zero speed may not be applicable to machine tools, but it is important in some applications, such as electric vehicles where it is the equivalent of slipping the clutch during a hill start.

Andrew

Unfortunately the question in the OP is unanswerable. Unless anyone has done long term reliability tests on a range of VFDs?

There seems to be some confusion regarding filtering and control schemes. Most smaller VFDs do not come with internal mains filters, and do not use power factor correction at the front end. The primary purpose of the mains filter is to reduce the current pulses taken from the mains. This reduces harmonic content and improves power factor. The mains filter is concerned with conducted emissions, not radiated. The problem of radiated emissions is mostly associated with the outputs of the VFD, where the PWM waveforms have fast edges and hence a high harmonic content. These radiated emissions are normally dealt with by shielding the output signals. I'm not going to discuss the ground at one end or ground at both ends controversy!

A simplified model of an induction motor says that torque is proportional to phase current. An induction motor will be designed to produce rated current at a rated voltage and a given frequency, aka base speed. In the UK rated voltage will normally be 415V phase to phase, in star, at 50Hz. As the frequency increases the currents will fall as no more voltage is available to drive extra current through the windings. Although current, and torque, fall the speed rises, so power stays constant above base speed. As the frequency decreases the rated voltage will drive more current through the windings. In the longer term this can lead to increased resistive heating of the windings and ultimate failure. As the frequency decreases it is normal to decrease the voltage to keep the phase currents constant. Torque also stays constant below base speed.

End of Part 1 - in case I hit the wrong key and the whole lot disappears.

Andrew