One thing I don’t see mentioned much is why so many model-engineers still use inches. Most work to long-established designs dimensioned in Imperial because the original machines were, but further because the standard model scales are to binary fractions of feet and inches.
Sort of…. scaling locomotives by gauge gives…
3.5″ g = 1 / 16 very closely. Fine!
5″ g. = 1 / 11.3. Who?
7.25″ g. = 1 / 7.79. Nearly eigth-ish.
And so already introducing compromises. I do not know the “garden-scale” inch:mm arcanities.
Other models, such as traction-engines and aeroplane engines, are scaled easily to true binary fractions.
Either way the drawings are not only dimensioned in inches, but to “things-over-bits” of inches and frequently no datum corners. So even with decimally-calibrated machine-tool dials we still have to convert the stated sizes to 10s-based. Might just as well go metric and be done with it IF your machines are metric.
On the other hand I would expect clocks, scientific-instruments and other non-model projects being all-metric.
I am used to both Imperial and Metric (yes I know it’s sniffily called “SI” but the main units are Metric), so I consider purpose.
An accessory for an old, British-made machine-tool? Inches and BS fasteners for compatibility including the fewer spanners the better. I wish the new Myford-branded accessories followed suit!
Extending my commercially-made clamping-sets? Compatible threads.
My model steam-wagon? Ah… well. This project has dragged on for so long it has become cross-bred, so I will bring most of its fasteners to BS (historically comformable though not vital) and BA standards. The components are all in inches and machined to decimal-inch dimensions. The exceptions are a few bought parts like clevises.
Anything else? As appropriate / convenient, frequently metric.
.
It is possible to become too hung up on this: use whatever you find the easier for your knowledge and workshop, but bear in mind the sun is setting on the Imperial.
+++++++
A little wander into the sidings…. as a relief from debating turning mm on 8tpi screw lathes with 1/128ths dials
Fulmen says the Pascal works fine. Does it? In Physics perhaps.
The Pressure given by a Force of 1Newton acting over 1 Square-metre.
I find the Newton awkward anyway, but even without that, think with the Pascal the SI scheme matches pure physics more than physical use. It is absurdly small. So mm, fine. Litres, fine. kg, fine. Pa, err…
100 000 Pascals for ordinary air-pressure, so 600 000Pa (0.6MPa) for most miniature locomotives.
Right, what can we use the Pa for, sensibly?
Sound pressure-levels? Those are tiddly, even at dangerous levels.
Ah, the Pascal is much too big for that. Oh dear… divide it by a million and count micro-Pascals, (µPa).
Now we have very wide ranges of awkward powers of ten, so use multiples of logarithms-of-ratios and call them deciBels. Only, that needs a reference-level equating to 0dB. Not equal to 0µPa as that creates a division by 0. So the measured sound pressure level is quoted as ” x dB re [the reference-level in linear units]” .
And for sound pressure level in air, because it is us lot for whom sound-levels really matter, it is logical to set the reference as the softest a fully-healthy human ear can detect.
It is staggeringly tiny, a mere 20µPa…. one-five-thousand-millionth of standard atmospheric pressure.
That flexes the ear-drum by about the diameter of a hydrogen atom.
Marine acoustics including biology is quieter still: just 1µPa for its 0dB: -26dB re 20µPa.