I'd think that 10% is too much, I'd go for "just detectable with a good ruler", or about 0.010" which agrees generally with your latter figure. But it also depends on the nominal dimension so perhaps a bit less on a dimension as small as 3/16".

I'd ignore it; it's a hang over from the days when most people didn't have access to micrometers but relied on plain calipers and rules. Just make it to size, or maybe a thou or two smaller if the part needs to fit in a hole.

Andrew

Yes really you just want the length to be enough so that when the nut is done up you just want enough clearance so the part on the shaft is not pinched so 1% or 1-2 thou would be about right. You do need to take the two mating parts into account to give the "fit" some context be it a shoulder as you have now mentioned or a shaft in a hole as you would want the shaft full for a press fit or the hole bare.

Although not mentioned in recent mechanical design any old ME article, books or plans are likely to have it and as Andrew hinted it goes back to a time when most hobby workshops would have had a steel rule, home made scribing block and a pair of firm leg callipers so little point is specifying 0.1885" when there was no way to measure that exactly.

The terms are well known in carpentry and joinery, the difference from nominal normally being taken as 1/64".

Edited By ega on 14/10/2022 14:00:39

I learned all my woodworking skills from my father and we used to say "line in" or "line out" to trim to the inside or the outside of the pencil line.

William.

I've never come across them either. The terms don't seem to me much use, as the actual value of the dimension would depend on the corresponding dimension of the mating part, which would complicate detailing considerably if there are multiple instances on a GA and it isn't instantly obvious which is supposed to be fitting what.

The Tool Design office I used to work in was in process of transition from Imperial to Metric standards – as indeed was the product – and in the Imperial case I *tended* to use the spectrum of fits from 'Loose Running' to 'Heavy Drive' with the mating Detail No. specified; whilst in Metric it would *usually* be ISO Hole Basis fits.

*My memory might be imperfect and I can't swear to've been totally consistent

It could get quite complicated sometimes when the component to be produced had Geometric Tolerances specified in Baroque detail. And I have to admit that I don't think I've seen any such since.

I have just been a-googling for a couple of minutes, and found this

.

.

The reference on Page 427 is particularly interesting [Mr Greenly, no less]

… does anyone have that 1904 issue to hand ?

MichaelG.

.

Edit: __ the image quality is Google’ problem, not mine !

.

Edited By Michael Gilligan on 14/10/2022 15:50:31

Inspector Meticulous would probably have something of the barrister as well as engineer in him when examining that paragraph.

My view is best avoid these terms.

Examine what the parts actually do, and work as closely as you can to achieve that. If a matter of fits in holes, where critical, we would not go far wrong by applying trade practice as outlined in, say, the Zeus Book.

We are after all, making either workshop tooling that has to be right to give correct results; or replicating old industrial products. So although most of us probably don't have Rolls-Royce Aerospace machine-tools and metrology standards in our sheds, I regard the arbiter as, "What would the trade do, and how closely can I match that?"

Excellent … Thank you, Nick

MichaelG.

.

Edit: __ Now, of course, I’m desperate to know what Oakham Splice wrote

… and, presumably, what Mr Greenly wrote before that.

Edited By Michael Gilligan on 14/10/2022 18:31:22

Thanks for trying, Nick … much appreciated

MichaelG.

In the 70’s when I worked in a contract drawing office some local machine tool companies were still using there own draughting standards. They not only had there own unique standard parts but also there own tolerances. Being UK based they probably used something based on the Newal tolerance system but then developed it based on there experience and the machines used. Remember in those days many people stayed with an employer as could be seen by the number of names on the 25 and 50 year employment boards. A company I went to later had about a dozen or more boards of 50 year personnel. So many of these people would be used to the terms on a drawing and wouldn’t need it written down just inferred. Of course all that is gone now and I doubt we will ever see a fifty year employee, I only managed just short of 25 before redundancy struck for the third time. Once British standard limits and fits took over standardisation became the norm and the old ways were forgotten along with the people who used them.