Accuracy of Hand Drilled holes

[Neil WyattModerator @neilwyatt]

Another thread has wandered from this question, for unsurprising reasons.

Can I ask more specifically, if you were marking out a circle out small (about 0.4mm) holes and then drilling them by hand, how accurately could you place them?

0.2mm? 0.1mm?

Thanks

Neil

[Neil WyattModerator @neilwyatt]

[Dave HalfordParticipant @davehalford22513]

That's an eye sight dependant / age question.

My daughter at 18 could be spot on easily, at 40 she could not as glass lenses had not been invented, or had they?

Edited By Dave Halford on 06/10/2021 10:09:07

[Neil WyattModerator @neilwyatt]

Posted by Dave Halford on 06/10/2021 10:06:58:

That's an eye sight dependant / age question.

My daughter at 18 could be spot on easily, at 40 she could not as glass lenses had not been invented, or had they?

Edited By Dave Halford on 06/10/2021 10:09:07

I'm assuming a skilled worker with no physical impediments.

So what is 'spot on'?

Neil.

[larry phelan 1Participant @larryphelan1]

I seem to remember Sparey commented on this in his book.

[Mick B1Participant @mickb1]

Posted by Dave Halford on 06/10/2021 10:06:58:

That's an eye sight dependant / age question.

My daughter at 18 could be spot on easily, at 40 she could not as glass lenses had not been invented, or had they?

Edited By Dave Halford on 06/10/2021 10:09:07

Also fatigue-dependent. On some recent work I wanted to reach limits like Neil's suggestions and *think* I succeeded, but there were only 4 holes to do, and bigger ones at that. And if I'd failed, the job could stand me opening the holes up a bit. I don't think I could have remained that accurate for more than perhaps 8 or 10 at a time, without a long coffee break between sessions. When I was younger, I might have done twice as many at a time, but eyes get tired at any age.

I was using a 3.5x magnifier – perhaps comparable with the Nimrud crystal from ancient Assyria held in the British Museum.

Then there's the question of how do we know the actual accuracy we've achieved, without spending more time and effort measuring results than we did on doing the job. Perhaps some of us might be prepared to invest that sort of effort, but in most cases I'd find it hard to justify.

Edited By Mick B1 on 06/10/2021 10:59:16

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Neil Wyatt on 06/10/2021 10:37:26:
.

I'm assuming a skilled worker with no physical impediments.

.

Then put-aside the measurements for a while, and consider the hole diameter as a ‘unit’

Would your idealised worker be satisfied with an placement error of +/- 50% ?

… The ‘skilled worker’ will find a way to improve upon that. [*]

MichaelG.

.

[*] __ Quoting from the reference to John Bird, on the other thread:

Bird prepared an 8-foot scale accurate with a vernier to 1/1000 of an inch. Few of his contemporaries could have made such a rule, and this was just to set the openings of his compasses.

Edited By Michael Gilligan on 06/10/2021 11:38:25

[Bill Davies 2Participant @billdavies2]

I make no claim for my ability to achieve such accuracy, but my instructor taught us to mark out perpendicular lines, pick up intersections and lightly mark with a sharp prick punch (60° centre punch), then using an eyeglass, 'nudging' the position until properly located, then apply a heavier tap. We used a small centre drill and opened up and reamed in the usual way. He claimed, with care, 0.005".

I think Chapman's first volume of Workshop Technology also claims five thou. I've put my copies somewhere safe, and can't find them. There is a set of all three on Ebay for £27.99.

Machine Tool Operation ( [US textbooks] Burghardt, etc., 1953, vol 1) claims it is possible to set external calipers 'easily' to within 0.002-3" against an engineer's rules. In those days, they were accurately ruled and deeply etched.

By my time, vernier height gauges were universally used, and we had mills and jig boring machines for accurate work.

Bill

[JasonBModerator @jasonb]

A lot will depend on how you are marking out, if its a steel rule and scriber I doubt you will place the lines as well as if you were using a height gauge. Number and spacing of holes around the pcd would also affect things eg 4 holes easy to place 27 less so.

Then what about punching, some seem to go straight in with a centre punch, I would possibly use a dot punch first and inspect the mark and pull it over if needed but more often that not I use my Veritas optical punch which gives me a more accurate result.

And what about hand drilling are we using a pin chuck, archamedies drill, egg whisk hand drill or handheld cordless drill. A non ferrous material will also be easier to drill than steel so less chance of wandering. Then their is quality of drill bit.

All comes down to what the individual has available to them and their ability to use it.

Edited By JasonB on 06/10/2021 13:02:41

[Michael GilliganParticipant @michaelgilligan61133]

Sorry … I am not at liberty to share my downloaded copy; but you may have access, or be able to find this paper elsewhere:

“Ancient machine tools for the construction of the Antikythera Mechanism parts”

**LINK**

https://www.sciencedirect.com/science/article/abs/pii/S2212054818300353?via%3Dihub

It’s interesting reading, but not a definitive answer to Neil’s question.

MichaelG.

.

Edit: __ This page is hosted by academia.edu

… so you may not be comfortable visiting it:

https://independent.academia.edu/AVoulgaris

Edited By Michael Gilligan on 06/10/2021 13:55:57

[Ian PParticipant @ianp]

Posted by Neil Wyatt on 06/10/2021 09:51:12:

Another thread has wandered from this question, for unsurprising reasons.

Can I ask more specifically, if you were marking out a circle out small (about 0.4mm) holes and then drilling them by hand, how accurately could you place them?

0.2mm? 0.1mm?

Thanks

Neil

I'll stick my neck out and say think I could do 0.2mm if I was careful and took my time (on something like brass), would be far less confidant about 0.1mm but I can imagine that a craftsman with unlimited time and a suitable magnifying glass could achieve that by continuously monitoring the start of the hole and tweaking (most probably) a bow driven or archimedes type drill.

Ian P

[Rod RenshawParticipant @rodrenshaw28584]

The late Tubal Cain had something to say about this type of thing in his article on co-ordinate drilling in his book "Simple Workshop Devices." The holes were a little bigger than those Neil writes about but TC does list some actual results of experiments to see what accuracy was possible by different means.

Rod

[pgk pgkParticipant @pgkpgk17461]

If one needed accuracy and had a lot to do then the obvious answer is to cheat – fix your hand-drill at the radius with some sort of vertical arm on a simple hinge. if the sheet you drill through is thin and the arm is long enough then any curvature becomes insignificant. Fix your plate to be drilled on a large enough circular support with witness marks on it's circumference and again the error is down to how large that disc is – a spare wagon wheel?
It depends on how much time you are given and whether the penalty for error is crucifixion.
A long strip of velum wrapped round the wheel then measured and divided into my chosen divisions by repetitive dividing.. say divide by three twice to 9 then by 5 to 45 and halves to 360 – or perhaps they got the sums wrong and that's why it isn't 360?
I’d prefer to go round with a simple spade point drill first to create dimples if talking brass and our modern options of drill types aren't around

pgk

[Jon LawesParticipant @jonlawes51698]

I can't drill with that accuracy using a DRO on the mill…

[pgk pgkParticipant @pgkpgk17461]

An interesting article postulating that ancient craftsmen may have been myopic with natural 'magnification'.
Link
https://www.penn.museum/sites/expedition/close-work-without-magnifying-lenses/

Also consider that with magnification there are some folk capable of such good hand-eye coordination that they can engrave the lord's prayer on a pin-head. I wonder ho accurate such folk can be even without a lens?

pgk

[Martin ConnellyParticipant @martinconnelly55370]

When I was at work one of the fitters was very myopic. If he wanted to look closely at something his glasses went up on his head and he held the object about 50mm from his eye to look at it. Maybe in an earlier time he would have ended up as one of these craftsmen doing detailed work by hand.

On a different scale of accurate hole drilling, we had a large portable Asquith radial drill that was a bit worn. When it was being used to drill large holes accurately positioned the initial centre pop had four additional centre pops around it on a PCD equal to the hole diameter. As the point of the drill was creating the initial cone shaped hole the process was halted and the position was compared to the outside centre pops. Any off-centre positioning was corrected with a die grinder and a bit more was removed with the drill and the process repeated until the drill was correctly following the desired hole centreline. The resulting hole had the half pops around the edge evidencing the good positioning of the drilled hole. Magnetic based drills eventually pushed the Asquith to the side-lines and it was eventually disposed of.

Martin C

[pgk pgkParticipant @pgkpgk17461]

The other thing we forget is patience. Many ancient construction projects took centuries from concept to completion. Even in later history a daunting project like a cathedral took more than 250 years to build with a huge team and often double that. It hardly seems unreasonable to image an orrery being designed and built over a few years of fiddling and faffing – a bit like any of my projects…

Imagine a few apprentices tasked with the grunt work of sourcing and preparing the blanks, perhaps practising on wood or stetched canvas panels coated with resins? Making patterns on a cheaper easier worked substance for later transfer? Unless there is a source of documentation then most archaeology is speculation often extrapolated from prior speculation rather than fact.

pgk

[Mike PooleParticipant @mikepoole82104]

To hit 0.1mm it would a lucky shot for me, to consistently get that would need much careful working. Would using a pillar drill count as by hand? I had the luxury of short sight from my early teens but ageing has robbed me of this and I now miss it. Drilling 0.4mm by hand would soon have a pile of broken drills if I was in charge of the drill. The marking out will need to be very good if it is not going to contribute its own factor into the final hole position. If you are taking the job on Neil then good luck and rather you than me.

Mike

[Bill PudneyParticipant @billpudney37759]

During my apprenticeship we had to make loft plates, typically 1/8" al.alloy, with an accurate photographic "picture" of what was required, printed on the sheet. They were usually developments of folded/pressed structural members, some were small 6" x 6", some were long and with a complex shape, the biggest I saw was about 72" x 12". The outline had lines from memory 0.006" or 0.010" wide, holes were indicated by crossed lines, with a circle appropriately sized for the drill bush destined for that hole, usually either 3/32" or 1/8" bore, with an outside diameter of something like 5/16" or 3/8". The bushes were very accurate, I think that they were sintered. The outline had to be filed to (ideally) half the line thickness, and the drill bush ideally would show half the line thickness all the way round. This is a long winded way of saying that, in my yoof I regularly achieved +/-0.003" or +/-0.005".

cheers

Bill

[Neil WyattModerator @neilwyatt]

Posted by Bill Davies 2 on 06/10/2021 12:21:05:

I make no claim for my ability to achieve such accuracy, but my instructor taught us to mark out perpendicular lines, pick up intersections and lightly mark with a sharp prick punch (60° centre punch), then using an eyeglass, 'nudging' the position until properly located, then apply a heavier tap. We used a small centre drill and opened up and reamed in the usual way. He claimed, with care, 0.005".

I think Chapman's first volume of Workshop Technology also claims five thou. I've put my copies somewhere safe, and can't find them. There is a set of all three on Ebay for £27.99.

Machine Tool Operation ( [US textbooks] Burghardt, etc., 1953, vol 1) claims it is possible to set external calipers 'easily' to within 0.002-3" against an engineer's rules. In those days, they were accurately ruled and deeply etched.

By my time, vernier height gauges were universally used, and we had mills and jig boring machines for accurate work.

Bill

That's particularly useful, thanks Bill.

[Neil WyattModerator @neilwyatt]

Posted by Bill Pudney on 08/10/2021 09:52:00:

During my apprenticeship we had to make loft plates, typically 1/8" al.alloy, with an accurate photographic "picture" of what was required, printed on the sheet. They were usually developments of folded/pressed structural members, some were small 6" x 6", some were long and with a complex shape, the biggest I saw was about 72" x 12". The outline had lines from memory 0.006" or 0.010" wide, holes were indicated by crossed lines, with a circle appropriately sized for the drill bush destined for that hole, usually either 3/32" or 1/8" bore, with an outside diameter of something like 5/16" or 3/8". The bushes were very accurate, I think that they were sintered. The outline had to be filed to (ideally) half the line thickness, and the drill bush ideally would show half the line thickness all the way round. This is a long winded way of saying that, in my yoof I regularly achieved +/-0.003" or +/-0.005".

cheers

Bill

And so is this, from another Bill, thanks.

[old martParticipant @oldmart]

It depends mostly on how skilled you are, little tricks such as moving the start point of the hole before finally committing yourself, and many people cannot drill square to the surface to save their lives.

[Nicholas FarrParticipant @nicholasfarr14254]

Hi, it's a long while since I've hand drilled tiddly holes that had to be reasonably accurately distanced apart from each other, it was probably in the early 90's when I was still making my own PCB's and components like transistors and dual in line IC sockets etc. needed to be inserted, The little plastic hand drill in the photo below is what I used and I could often do up to forty or more holes before the drill bit broke, which used to cost about 15p each from a local tool store at that time.

I suppose one could say I was cheating on the marking out of the hole positions though, as I used these etch resistant transfers that Maplin electronics used to sell, which I found were very successful.

Regards Nick.

[Kiwi BlokeParticipant @kiwibloke62605]

Marking out and drilling by hand? I wouldn't even try.

If this refers to the Antikythera mechanism, why assume it was done by such primitive means? The designer of the thing knew more than a bit about maths and geared mechanisms. Therefore I'd expect a dividing engine would be known technology, or at least imaginable to the designer, and a drill bush in a stand would facilitate accurate drilling. Those folk were SMART!

[Neil WyattModerator @neilwyatt]

Smart enough for KISS to apply.

The mechanism itself is, in essence, a dividing engine with many ratios built in. The makers would have been well aware of its accuracies and inaccuracies, including knowing that angular errors in the gears would have been trivial and cyclical – only errors due to gear ratio approximations are cumulative.

They needed a circle of 'x' holes to step the mechanism around in 'x' steps per rotation.. Micron accuracy is not required, so there is no reason for them to go chasing it.

To me the obvious solution is scribe a ring, use a marked strip wrapped around a circumference as an index. Scribe or punch each hole, then drill by hand.

Similarly, gears only had to be accurate enough to run and being hand finished could be eased if there were any tight spots.

[Author]

Posts