New Micrometers

Reading inaccuracy caused by a reading circuit problem giving a believable false reading. A mechanical instrument can't present such a reading. Colin

Surely mechanical instruments give believable false readings whenever the screw is worn, anvils damaged, the clutch is iffy, or the frame is bent. My main objection to second-hand micrometers that they are all believable!

I twitch when chaps describe their latest bargains as having 'reputable brand-names' or being 'top-quality', 'high-grade', 'decent', 'good', 'spot-on', 'very accurate' or 'as new'. These subjective terms are inappropriate to an instrument used to take accurate precise measurements. They imply the owner is an optimist rather than an engineer, and certainly isn't a metrologist!

Micrometers have a specification proved by checking them periodically against a reliable standard gauge and recording the results on a calibration certificate. The standard isn't a pre-loved set of Jo-blocks, even if they do look tip-top. Proof and evidence please, not unverified opinion.

I argue digital displays are less likely to mislead when they go wrong because the malfunction is usually obvious: blanks, zeros, eights, negatives and jumping digits etc. In comparison, an analogue display always looks good, even when the instrument is out of tolerance.

Dave

If you rely on using the clutch to take readings it is easy to inadvertently include a speck of swarf as it is not felt ; it is my opinion that finger touch on the thimble is more likely to show up such an event . Although I confess I've never experienced it on a micrometer, I have seen a clock digital display with a number giving a wrong figure due ,presumably, to the connecting matrix becoming faulty and losing a part of the digit. Colin

My fear with a digital micrometer is the same as with a digital calliper, accidentally coming out of "Abs" mode and not noticing. To try and avoid this I set the "Inc" zero well away from the zero point which hopefully would make this error obvious.

Brian G

This graph puts an interesting perspective on accuracy and the cost of achieving it with various machine tools.

The graph dates from 1991 and doesn't reflect improvements made at the high-accuracy end: it's not unusual for a CNC Machine Centre to do rather better than 0.025mm.  Broadly right though.

The cost of accuracy increases linearly up to a 'knee', beyond which the cost multiplies. The graph shows the maximum repeatable accuracy rather than the very best that can be achieved by a skilled operator taking his time and accepting a lot of rejects!

The graph divides in the middle at 0.025mm, which is a thou. 0.025mm sets a boundary between low and high accuracy, and it's about the best that can be expected of Shaping, Planing, Horizontal Boring, Milling machines and Capstan/Turret lathes.

Planing and Shaping are expensive and not very accurate. Although they can be pushed to the 0.025 limit, they perform most economically when less than 0.1mm accuracy is needed. Their high cost compared with milling probably explains why these machines have become unusual. My guess is they are too slow.

Milling gets to 0.1mm fairly easily, but pushing up to 0.025/thou triples the price. It's unlikely that a home-workshop mill can achieve better than 0.025mm accuracy. The same limitation applies to Turret and Capstan lathes: they're good for cheap production and can achieve about the same accuracy as a mill.

High-accuracy requires grinding, including honing, lapping, and hand-lapping. Precision grinding is rare in a home workshop, but industry use it on a huge scale.

Lathe performance as shown on the graph is interesting because it straddles the 0.025mm boundary, getting to about 0.02mm/0.0008" before the knee kicks in, with a practical limit of about 0.013mm/0.0005". Lathes are good performers, capable of reasonable accuracy without costing a fortune. Likewise, milling machines are attractive - reasonable accuracy at low cost. Model engineers don't do much precision grinding because the machines and process is relatively expensive.

That 0.025mm boundary is important in the context of micrometers because it pretty much defines the practical accuracy to which we work. It's about a thou, and doing better than ±0.0005" is really hard. I doubt many home workshops are capable of it, whatever micrometer they have.

Most of us use a different technique I think? I 'fit' parts together, using one to gauge the other, without worrying about exact dimensions. For example rather than bore an accurate 20mm hole and turn a 19.97mm shaft to fit it, I bore a hole of about 20mm, and turn the shaft to slide comfortably into it. No need for a micrometer of any sort, let alone one that does tenths! Fitting is too slow and expensive for production work and spares need to be interchangeable but I don't care!

Does anyone really apply the American System of Manufacturing at home?

Dave

Edited By SillyOldDuffer on 05/12/2021 17:47:21

What I do like about mechanical micrometers is their are no batteries to leak and fail the tool. In saying that, I recently did buy a digital depth micrometer, for the purpose of using it like a comparator. I don't think it matters all that much , if electronic or mechanical, as long as you can make the parts you want to. Gauge block sets are in the very affordable range now, with many coming with a certificate of accuracy for each block. So with a good set of gauge blocks and repeatable measuring equipment, new cutting tool technologies, now more than ever, allows home users to being able to easily make interchangeable parts.