Metric or Imperial, Fractions or Decimals

[TerrydParticipant @terryd72465]

Hi John,

 

All I pointed out was that Mathematically fractions are exact numbers, inferring calculations using fractional numbers i.e. nothing to do with machines or measuring. Just because computers and calculators use digital and decimal systems which are often at best close approximations does not alter the fact I said nothing about fractional measures except that we used to use them for large structural projects. I wish folks would read posts properly and fully so that they can understand exactly what is meant before commenting.

 

By the way, try making a set of hidden mitre dovetails with an axe.

 

Regards

 

Terry

Edited By Terryd on 23/01/2011 11:14:23

[Nigel Graham 2Participant @nigelgraham2]

The practice cited in the OP, of both fractions and decimals of inches, was also once common in industry.

I would advocate that consistency and how the dimensions are given, are what really matter though.

If using inches, you could use 3-place decimals for machined parts but 2 where less critical, but given that it's probably much more common now to use machine-tools for most of our workshop operations, all-decimal anyway would make sense.

I would advocate too, thinking how the part is to be made before dimensioning it, to avoid the builder having to re-calculate datum faces and the like. Not everyone has digital read-outs on their machines, let alone NC machines, but even with these it is still worth thoughtful datum locations to minimise movements and reversals.

'

As for Pi… Mathematical Pi is still published!

[Brian HParticipant @brianh50089]

With regards to machines, with the advent of cheap DROs it really doesn't matter. I normally work in inches because I make models of 18th & 19th century items and my 'newest' machine is a metric Senior Major ELT but it came fitted with Mitutoyo DRO so no problems.

Brian

[old martParticipant @oldmart]

Using either type of decimal measurements is fine, but try to avoid fractions.

[Roderick JenkinsParticipant @roderickjenkins93242]

Guys,

Thank you for your input. You've resurrected my very first forum post from 10 years ago

Cheers and stay well,

Rod

[Howard LewisParticipant @howardlewis46836]

When dimensioning drawings, in what ever units you choose, it is better to use co ordinate dimensions rather than "point to point", since any build up of tolerances or errors should not then be cumulative.

If you consistently make an error, the effect is likely to be less than dimensioning from one point to another, (which will make the error cumulative ). It is amazing how the odd thou , each time, can build up into an unacceptable discrepancy.

Co ordinates is the way that Industry has done things for many years.

Howard

[Mick B1Participant @mickb1]

Posted by John Olsen on 22/01/2011 23:50:07:

…

I alway though the arguments in vavour of first angle were pretty specious…

 

…

regards

John

 

So do I. If I got a first angle GA of the Titanic, it seems crazy I gotta hike to the bow if I wanna look at the propellers…

[Ian PParticipant @ianp]

Posted by John Stevenson on 23/01/2011 01:45:38:

Fractions?

 

I have yet to see any machine tool with the dials graduated in fractions.

 

Wood working machines don't count, we are talking accuracy here not axe strokes.

 

John S.

My hero!

Ian P

[Martin KyteParticipant @martinkyte99762]

Posted by John Stevenson on 23/01/2011 01:45:38:

Fractions?

 

I have yet to see any machine tool with the dials graduated in fractions.

 

Wood working machines don't count, we are talking accuracy here not axe strokes.

 

John S.

My Myford is graduated in 1/1000" with additional graduations every 5/1000" and 1/100" the 1/100" marks being stamped in 10's. Not a decimal point in sight. What's that if not fractions.

regards Martin

[Paul LousickParticipant @paullousick59116]

My metric mill has dials which are in fractions and graduated in 1/1000 metres but like imperial machines with graduations in 1/1000" it is using a denominator with 10 as the base number. (decimal system).

I started my career as an engineering draftsman and remember calculating lengths in fractions before the invention of calculators. A real pain in the butt ! (Try adding up: 3' 5. 37/64" + 1' 2. 17/32" + 4' 11. 9/16", )

Paul

[Bill PudneyParticipant @billpudney37759]

Posted by John Stevenson on 23/01/2011 01:45:38:

Fractions?

 

I have yet to see any machine tool with the dials graduated in fractions.

 

Wood working machines don't count, we are talking accuracy here not axe strokes.

 

John S.

Trust the late John S to sum it up in a couple of sentences!!

cheers

Bill

[Ady1Participant @ady1]

Perhaps we should invent a new measurement which converts imperial/metric with perfect precision

widgets

[Michael GilliganParticipant @michaelgilligan61133]

Just returning to Rod’s original question … I recall posting this in 2013

___________

Just an aside …

It was quite common, until at least the mid-1960s, for American draughtsmen to work in 1/128ths of an inch.

I had hours of fun trying to understand all the strange dimensions on the drawings for a Polaroid 5×4 film holder … until I realised that it had been designed in 1/128ths and then later "modernised" to decimals. Once I had reverse-engineered the "rounding", the design made sense.
___________
.

The moral [if any] being that perhaps drawings should be dimensioned, or at least co-dimensioned in the system of units appropriate to the original design. …

MichaelG.

.

Ref: https://www.model-engineer.co.uk/forums/postings.asp?th=90017

Edited By Michael Gilligan on 23/05/2021 08:08:01

[J HancockParticipant @jhancock95746]

Or , as my old woodwork teacher used to say , " Engineers have to work to within ' a thou ', carpenters have to

be exact."

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Ady1 on 23/05/2021 08:03:23:

Perhaps we should invent a new measurement which converts imperial/metric with perfect precision

widgets

.

We did … when we declared the inch to be exactly 25.4mm

MichaelG.

.

http://metricationmatters.com/docs/WhichInch.pdf

 

Edited By Michael Gilligan on 23/05/2021 09:57:52

[Nigel Graham 2Participant @nigelgraham2]

Martin –

No-one said decimals are not fractions, but the thread's colloquial language somehow suggests that fractions exist only in the vulgar form!

'

No, of course there isn't a decimal point on a machine handwheel that counts entire thousandths or an inch, or fiftieths of a millimetre… but I've still to decipher the faint lines on my Denbigh H4 milling-machine's dial, on a feed-screw of one-sixth inch lead! Very few regular decimal dimensions match sexagecimals closely.

'

I've worked with Imperial drawings dimensioned in both in vuglar binary and decimal fractions. The tolerances were given as +/- 1/64", and 0.0xx" , respectively; but any machining and the inspection was all decimal.

'

Michael –

1/128" … Strange limits. I wonder why they didn't go straight to thous or at least hundredths. I once had a steel rule on which one of the inches was graduated in one-hundred-and-twenty-eigths – as if anyone could be expected to use it sensibly. I still have one of those transparent plastic rules intended for drawing-office and laboratory use, with a similar-triangles scale that "magnifies" titchy bits of inches to legibility. Never had to use it!

In line with your comment about matching system to original, I am used to Imperial and metric but am building my engine to one-third scale, in its ancestors' feet and inches. The drawings I make reflect this but with decimal equivalents (TurboCAD, presumably any CAD, uses only 10s-base, Imperial or metric). Even so I often use mm when measuring its physical steelwork to verify alignments or match non-machined parts directly rather than via drawings.

I used all-inches to design and build my workshop's travelling-hoist, but due to its metric fastenings, metric steel-stock and increasing complexity, I wished I'd gone all-metric. It would have been easier!

[Martin KyteParticipant @martinkyte99762]

I was being a little tongue in cheek Nigel. Everything less than unity is a fraction. As you intimate it's the coice of counting base that is less or more convenient in different settings. If you habitually divide things twelve is usefull as it has more common fractions which is why binary fractions work so well. It's also marginally less likely to make errors in systems that are not entirely homogenious. 234mm is easily confused with 243mm. 9 and 7/32 inches less so.

Specifying a drawing in fractions and thou also gives some indication of required tolerance.

I use whatever is convenient.

regrds Martin

[Paul LousickParticipant @paullousick59116]

Why does "Specifying a drawing in fractions and thou also gives some indication of required tolerance" differ from giving dimensions in mm ?

Paul.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Nigel Graham 2 on 23/05/2021 11:28:42:

'

Michael –

1/128" … Strange limits. I wonder why they didn't go straight to thous or at least hundredths. […]

.

The original 5×4 interface was defined by Graflex … and dates it back to the days when camera bodies were machined from wood.

I can only assume that Polaroid’s design draughtsman continued with the ‘binary’ fractional system … and then it was decimalised for production.

MichaelG.

.

https://youtu.be/JlUWIR7b6CE

Edited By Michael Gilligan on 23/05/2021 12:59:46

[Howard LewisParticipant @howardlewis46836]

Another piece of hardly relevant information.

The makers of Graflex cameras, at one time employed a man to travel around USA buying up old pianos to provide the wood from which to make the cameras! Presumably because it would be well seasoned by then and less likely shrink or warp.

Howard

[Martin KyteParticipant @martinkyte99762]

Posted by Paul Lousick on 23/05/2021 12:35:21:

Why does "Specifying a drawing in fractions and thou also gives some indication of required tolerance" differ from giving dimensions in mm ?

Paul.

What I meant was it was inferred in the dimension. A 3/4" dimension of a hole away from an edge would infer it was fine to measure with a rule, scribe and dot punch then drill the hole. A 0.750" dimension would indicate something requireing greater accuracy such as a toolmakers button. Of course you can always specify explicitly tolerances and if you dimension drawings in mm only you would have to do it. It's just an observation not a suggestion that you must do it one way or another.

regards Martin

[Nick WheelerParticipant @nickwheeler]

Posted by Martin Kyte on 23/05/2021 12:18:29:

Specifying a drawing in fractions and thou also gives some indication of required tolerance.

doesn't a trailing zero do the same? 0.1550 would suggest more precision that 0.155.

Which removes the need to reconcile weird fractions and seemingly random decimal numbers used on the same part.

[Mark RandParticipant @markrand96270]

Umm. Surely on the drawing you always put a table showing words to the effect of:-

Untoleranced dimensions below 1/16" ± 0.002". Below 1" ± 0.05". Below 12" ± 0.01". above 12" ± 0.015". Tolerances are not to be cumulative.

[Martin KyteParticipant @martinkyte99762]

Nick and Mark

Of course it does and of course you can. I did say that. Merely pointing out something you get for free when specifying fractions. Probably more normal on a sketch than a fully worked up drawing but usefull none the less.

regards Martin

[Bill PudneyParticipant @billpudney37759]

Then of course, if you use a sensible system, like the metric system, life becomes so much easier for the draughtsman, the person who is working to the drawing, any subsequent inspection, and historians analysing the drawing in 50 years time.

As for tolerances, oh dear oh dear…………. As a former contract draughtsman who worked in many companies with just as many drawing office systems there was little or no uniformity regarding tolerancing. The more organised companies had a system sometimes unique to them which avoided inferring anything, everything was specified in black and white. The less well organised companies permitted just about anything. As far as I can see one of the few advantages of globalisation is that drawings have to be drawn to uniform standards, including methods of tolerancing, simply so that the drawing can be used just about anywhere in the World, to produce identical parts. Unlike days of yore when measurements were based on the length of Freds' arm, finger or big toe, and tolerances, if shown on a drawing at all seem to have been based on something fairly random.

cheers

Bill

[Author]

Posts