Metric Screws and Threads

[BazyleParticipant @bazyle]

I assume American designed models use UNC/UNF. Do they have modeller’s versions with different sized heads?

Now I come to think of it when I started work in ’77 for a UK Defence Contractor we had metric and UNx fasteners in stores but no Whit or BA. We used the UN for all jigs and fixtures but the actual product had to be all metric, even if a part was ordered from the USA. This ensured loads of NATO equipment was incompatible depending on which side of the Atlantic it came from. Once during tests in Florida we couldn’t refuel the UK aircraft on an American airbase because they had no compatible hoses. (but the local equivalent of B&Q sold plumbing parts spec’d as BSP)

[JasonBModerator @jasonb]

Yes, companies like American Model Engineering do fasteners with smaller hex, just as well too a sfrom what I have seen of UNC they come up even bigger across flats than a similar size standard metric

[Robin Dufton]

On Robin Dufton Said:

… It seems to be something that upsets people …

I obviously don’t qualify as “people” then !

Nothing you are wanting to do upsets me in the slightest.

MichaelG.

[Nigel Graham 2Participant @nigelgraham2]

I chose UNF (not UNC) based on using standard fastenings, but this was at large scales where faithfully reproducing hexagons to scale is less significant.

The main fastening size for most of the chassis, is 1/4″, with some 5/16″ and 2BA (3/16″ – ish).

A 1/4″ UNF hexagon is 7/16″ A/F. (0.437″ )

1/4″ BSF – 7/16″ A/F nom. The bolt I measured was nearer 0.440″

M6 : 10mm A/F nominally (0.390″ ) . I measured both a nut and bolt and neither was very accurate!

.

Now lets’ consider what happens if we scale them up.

I abandoned the six-inch scale attempt as an over-ambitious failure, and started again in a more sensible 4″ scale, but still using 1/4″ or M6 fastenings.

 

So 3 X 1/4 gives 3/4″ BSW in full size, which seems right, though perhaps the maximum, for the bolt.

However, the nuts and the bolt heads I am using give 1-5/16″ A/F by scale alone. Would that be right?

Let’s find the BSI specifications for “bright” nut and bolt hexagons, Revision dated November 1908. For bolt diameters the max A/F, with -0.005″ tolerance, in inches is:

1/4   0.525

5/16  0.600

3/8    0.71

1/2    0.920

5/8    1.100

3/4     1.300

1       1.670

So sayeth the British Standards Institute, quoted in (Spooner, 1913). I read “bright” as all-machined rather than forged and threaded.

Therefore if the originals were 3/4″ BSW,

The 1/3 scale model’s 1/4″ bolt hexagons should be (1.3 / 3) = 0.433 r. A/F to impress Inspector Meticulous.

So 1/4″ UNF (0.437″ AF) gives an error of 0.004″ , lost in tolerances.

and 1/4″ BSF bolts to modern proportions are 0.440 – 0.433 = 0.007″ up. (give or take tolerances)

ISO-M6 Coarse? 0.433 – 0.390 = 0.043″. Over 1/32″ down – but frankly, especially when you’ve painted everything, is anyone going to notice even that?

.

Suppose the original was held together with 5/8″ BSW bolts (feasible). The scale shank diameter should be 0.208″, and the hexagons 0.367″ A/F.

M6? 0.390″ (0.023″ too big A/F but probably not excessively. The thread is 0.025″ too large but hidden.

M5? 0.313. (0.054″ too small – nearly 1/16″ error so excessive)

2BA? 0.372″ (0.005″ large A/F – acceptable) (Though 0.180″ thread dia, so nearly 1/32″ thin.)

 

Note that M5 / 2BA difference. These are often taken as equivalent irrespective of scale, but though the thread diameters are close the hexagons certainly are not, for true scale appearance!

 

So 2BA is the nearest scale size for 5/8″ BSW bolts made to the 1908 standard.

Only on my model wagon’s chassis, they would look far too small! It is of 50 X 25 X 25 mm rolled steel channel, right for the only overall size quoted in the contemporary trade reviews, but probably over-scale thickness.

 

These differences may matter in fine-scale replicating of the more delicate, exposed parts like motion-work and boiler fittings. Otherwise we may become over-worried about errors hiding in plain sight and too small to see.

Clearly we do not want obvious over-scales like nut corners overhanging the part edge, or not permitting room for the spanner. Neither do we want fastenings that look, or indeed are, too weak – but strength is a function of thread core area not head size. Nevertheless I think we can go a bit too far and make life unduly difficult for ourselves.

Allowing for some builders being unable to obtain certain ranges of fastenings, it may be that for closest fidelity we should not insist on this or that thread series by convention or modernity necessarily, but on the series that gives the best results. The more important criterion really is consistency throughout the model, but this might also be a prime application of “if it looks right it is right”.

[JasonBModerator @jasonb]

It is not just the hex size that needs to be considered but also the height.

In the smaller scales lets take 2″ as that is what the OP mentioned  for his traction engine height of a pre war3/4″ whit nut is 0.75″.

If we take M3 as the nearest to 1/6th of 3/4″ then the nut thickness is 2.4 – 2.15, lets say 2.25mm average.

6 x 2.25 is 13.5mm which is a lot less than 19.05mm

Add to that the soft edges and double chamfers of mass produced nuts and they just don’t look right.

 

[JasonB]

On JasonB Said:

Add to that the soft edges and double chamfers of mass produced nuts and they just don’t look right.

Which is why I am making almost all the nuts, bolts and studs for my engines:

The most common sizes are 1/4″ BSF and 2BA. The 1/4″ BSF nuts are mostly 1/4″ high. There are a few places where that doesn’t look right, or the standard nuts will not fit. There are also a number of home made M6 fittings for space saving, or to fit with externally supplied parts. In the internals, where they won’t be seen, there are a number of commercial M6 SHCS fittings. There are also some small home made metric screws, down to M2. The few small BA screws (down to 12BA) are mostly commercial. Not worth setting up the repetition lathe for very small quantities.

My engines are a real mix of threads, as well as the above there are some UNF threads and a lot of the boiler fittings are 3/8″ BSP and 32/40tpi ME threads. No hang ups about metric or imperial, I just use what is appropriate or convenient.

I’m not sure what the point, or question, was in the OP as the replies were clearly not welcome.

Andrew

[Nigel Graham 2Participant @nigelgraham2]

The OP is querying close BA / Metric matches, and the feasibility of making them; though why you think the replies are “clearly not welcome” I am not clear.

Essentially our using either BA or Metric fastenings to represent the full-size machine’s BSW or BSF – and then what age of BSW – will always be a compromise between function, appearance and scale fidelity unless we go to the lengths of making the fastenings ourselves.

Robin’s question reads to me as on the feasibility of making ISO-metric versions of BA, though, having apparently started to mix the two systems in one project. (Errr, as I have!)

If we do make the nuts and bolts to true scale they are non-standard, so the thread type is not critical, within reason, and can be as we consider best for our models and tooling.

..

I wonder what happens if we replicate to scale, a machine made originally to metric dimensions with metric bolts? Would modern ISO-M Coarse or Fine fasteners be true to scale?

[JasonBModerator @jasonb]

I think the OP’s biggest problem is he is wanting to substiture BA with metric rather than starting with what the original sizes may have been and working back from that.

Someone who originally designed what he is making will have made compromises to arrive at the nearest BA and then compromising again going from BA to Metric will risk drifting further from the original.

If you are going to use predominantly metric threads then you also need to start looking at any shafts and rods that are threaded too so that they are the right nominal size and then what those shafts pass through so may as well just metricate the whole thing. Both Tooling and materials are often easier or cheaper to get hold of which is another advantage.

 

[Nigel Graham 2]

On Nigel Graham 2 Said:

… though why you think the replies are “clearly not welcome” I am not clear.

…

I would never presume to answer for Andrew, but I think Robin’s closing remark on page_1 indicates that some of the replies have strayed too far from what he was seeking.

MichaelG.

.

… It seems to be something that upsets people so I’ll go back to building my models in private and running them around my garden.

[Michael Gilligan]

On Michael Gilligan Said:

On Nigel Graham 2 Said:

… though why you think the replies are “clearly not welcome” I am not clear.

…

I would never presume to answer for Andrew, but I think Robin’s closing remark on page_1 indicates that some of the replies have strayed too far from what he was seeking.

MichaelG.

.

… It seems to be something that upsets people so I’ll go back to building my models in private and running them around my garden.

Remember the scene in Cool Hand Luke, where after beating the hero to the ground, the chief warder says ‘what we have here is a failure to communicate…”

Though Robin knows exactly what he wants and why, the rest of us are confused.  Or at least I am!  It’s because the answer depends on what Robin is trying to achieve.   Reading between the lines, and I could be entirely wrong, I guess Robin is metricating the BA fasteners specified in an old design, perhaps to suit his metric workshop, or maybe because metric fasteners are easier to source than BA.   In that context, it’s not clear how important the aesthetics are to Robin, and that generates debate.   Robin expected a simple answer, and there isn’t one.

Robin asked: Making the 4, 6 and 7BA replacement screws is getting a bit of a drag. Would there be enough people interested in switching over to screws these sizes if I asked around for someone to make batches of them.

The simple answer I think is No, but not for a single reason.  It depends!  My workshop has no need for 4, 6 and 7BA near match metric fasteners because I mostly do original experimental work where scale proportions and other aesthetic considerations don’t matter. The need doesn’t arise because I don’t work from Imperial, US or BA plans.  At the other end of the spectrum, Andrew Johnston is unlikely to need them either because the correct appearance of his traction engines, scaled from Imperial prototypes, is important, and he doesn’t compromise when it matters.   I guess Robin is in the middle somewhere, building a metricated BA-based engine for his private garden railway that looks good to him without having to satisfy Inspector Meticulous.  There must be other people in the same boat, but probably not enough of them to create a demand for special screws in BA similar sizes.

Worth mentioning that none of the standard fastener systems were designed to make life easy for scale modellers.  If smaller nuts and bolts look OK, it’s a coincidence.  Specifying BA in a model build was itself a compromise, a way of creating a reasonable look with off-the-shelf fasteners, and avoiding the need for modellers to make their own.

Or maybe Robin is a pioneer, one of the first to smack into a modern problem!  In the past, it wasn’t difficult to source obsolete fasteners, making it possible to follow old plans exactly.  Although common Imperial & BA sizes are still available, time is remorselessly reducing their availability, especially of unusual sizes and head-styles.  Could be Robin has found he can’t get BA in the sizes needed, been forced to metricate, and discovered it ain’t straightforward.   If so, the best answer in the long run is for popular classics to be redesigned in metric from the ground-up, so that new builders aren’t faced with demands for materials and parts in unobtainable Imperial sizes.

Dave

 

[SillyOldDuffer]

On SillyOldDuffer Said:

 

Robin asked: Making the 4, 6 and 7BA replacement screws is getting a bit of a drag. Would there be enough people interested in switching over to screws these sizes if I asked around for someone to make batches of them.

The only real sticking point is that similar diameters to 4BA are not readily available in the metric system in the forms we want them in.

4BA at 3.6mm dia would best be served by using M3.5 but I have not seen this being offered in model fixings. I have suggested picking what suits best from either M3 or M4 which has worked for me.

6BA at 2.8mm dia falls part way between M2.5 and M3 so you would have to pick one or the other as no taps or dies are available as an exact match but off the shelf model fixings are available in M2.5 and M3 so no need to get somebody else to machine special batches

7BA at 2.49mm is a no brainer and you can use M2.5 for that, Again model fixings are available so no need for batches.

Of the others

0BA at 6.0mm is another no brainer to use m6

1BA seldom seen on model drawings so forget about it but if you do find it holding on a Unicorn’s horn then M5 is very close

2BA Which is a commonly seen thread as it will work on 3/16″ stock can make use of M5

3BA at 4.1mm is not far from M4

5BA at 3.2mm again a common thread on models a sit works on 1/8″ stock is usually substituted for M3

8BA at 2.2mm use M2

9BA not common on models but at 1.9mm you can use M2

10BA at 1.7mm is best served by M1.6 and if you want nominal material to threda use 1/16″ stock.

So if batches were being ordered then really only M3.5 is needed possibly with a 6.0mm hex that should be readily available hex size. Though there is still the question of getting hold of 3.5mm stock for studs and even the 9/64″ option is not easy to find, I think just EKP stock it

Or do as I have and just use M3 or M4 and don’t worry about the gap between them.

J

 

PS I have avoided the obscure metric diameters before someone chips in about them as the likes of M1.7, 1.8, 2.2 and 2.3 are specials and stock to cut then from for studs etc would most likely need turning down from larger

[JasonB]

On JasonB Said

The only real sticking point is that similar diameters to 4BA are not readily available in the metric system in the forms we want them in.4BA at 3.6mm dia would best be served by using M3.5 but I have not seen this being offered in model fixings. I have suggested picking what suits best from either M3 or M4 which has worked for me. […]

So if batches were being ordered then really only M3.5 is needed possibly with a 6.0mm hex that should be readily available hex size. Though there is still the question of getting hold of 3.5mm stock for studs …

Please forgive me if I appear thick …

To me, that all sits quite nicely with my mention of the wide availability of pattress screws.

The Electrical trade grabbed the M3.5 with open arms … but didn’t have the need for ‘Model’ heads.

So why not run a big batch of nuts, suitably proportioned … to be used either as functional nuts, or fixed onto a length of screw to make ‘bolts’

The screws are dirt cheap, and easy enough to trim to length.

MichaelG.

.

.

[admission] making proper studs, incorporating an unthreaded section, would remain a challenge.

[JasonBModerator @jasonb]

I think I would rather get hex head screws/bolts made or make my own. Looking at my usual electrical wholesaler their M3.5 x 25 screws are 17p each add similar for a model nut so lets say 30p each. A hex bolt would be about 20p based on cost of M3 and M4.

Add to the fact most switch plate screws are plated you would not be able to silver solder on a nut without problems and the faff of cleaning them up.

I expect you can get cheaper m3.5 screws but just like cheap studding the quality of the thread can be suspect.

Personally I’m not so bothered about bolts/screws as they make up less than 5% of fixings on a typical model that I make. Of those few bolts the ones mostly used on steam subjects are “fitted” bolts in things like eccentric straps so would not want to use fully threaded shanks anyway. Olber Model IC engines may want Fillister head screws and some of the open crank engines square head so I’d not have much call for using electrical screws.

[Nigel Graham 2Participant @nigelgraham2]

There are three aspects here, assuming we wish to use metric fastenings:

1) Equivalence to the BA or other standards we are replacing, for function: i.e. is the replacement just as strong (assuming sufficiently similar types and grades of material).

2) Need for plain shanks whose lengths may have to be correct for the application so not necessarily obtainable anyway.

3) Scale proportions: the bolt or stud shank and thread are hidden; but the heads and nuts may be very visible; and may need be square or of some other form like a coach-bolt.

 

Jason has given us a summary of Point 1): functional replacement of BA by Metric. Similar can be applied for larger sizes:

1/4 ” BS or UN can often be replaced by M6 (slightly smaller)

5/16″ -> M8 (very close diameter match)

3/8 -> M10 (the 10mm thread is almost 0.5mm larger diameter.

etc.

Fastenings of larger diameters are probably unusual on models, but may occur on miniature-railway rolling-stock and structures, workshop equipment, etc., and here we may as well use metric because these are not “scale models”.

.

So what of Points 2) and 3). Really, these are where we need consider making the fastenings, especially with plain shanks (e.g. “fitted” bolts) or scale fidelity, or both. Then the thread type is whatever we want, within reason; and we might even need use non-standard combinations of size and thread.

Making very long bolts (plain shanks longer than the threaded portion) from solid is a bit extravagant, so assess if it will be more economical to make a stud – i.e. with a long plain section – and permanently secure an appropriately-proportioned nut on one end as the “head”.

E.g.: two long, 5/16″ BSF T-bolts I made recently, from plain rod. Head screwed tightly to the thread run-out with a dash of locking compound; the assembly cured overnight, then the head and nub of projecting stem faced to finished thickness.

I have some very long 2BA bolts that were commercially made, probably for holding motor end-plates together. Very close inspection shows the same technique though brazed not glued, before the assembled bolts were plated.

.

Somewhere above is a comment apparently about the awkwardness of making studs.

One, perhaps the easiest, way is to put one thread on each, the longer if they are unsymmetrical, first.

For the second end, hold each in a piece of rod drilled and tapped through, with a locking screw entered from the far end. This is then held in the lathe chuck or collet with a back-stop as precaution and for consistency, to trim and thread the second end.

If stud is long relative to its diameter, drill the outer end of the jig to close clearance diameter for sufficient depth to support the shank.

The arrangement is best sized so the embryo stud is not driven onto to its thread run-out.

 

[duncan webster 1Participant @duncanwebster1]

If using Nigel’s method of gripping studs, it can be advantageous to either use a left hand thread on the plug at the far end to eliminate the possibility of it working out under the torque from the die, or much simpler, loctite it in.

[Michael Gilligan]

On Michael Gilligan Said:.

[admission] making proper studs, incorporating an unthreaded section, would remain a challenge.

Not sure why; studs are fairly simple, albeit it needing two operations. First operation is to round the end of rod, thread the top of the stud and part off. Then the embryo stud is screwed into a threaded holder with adjustable back stop and the bottom of the stud is threaded, leaving a plain portion:

I am making all the studs for my engines, mostly 5/16″ and 1/4″ BSF, M6 and 2BA:

The large studs with a plain end are for holding the cylinder down to the boiler. Until the cylinder is fully machined I don’t know exactly how long the plain section needs to be.

Andrew

[Andrew Johnston]

On Andrew Johnston Said:

On Michael Gilligan Said:.

[admission] making proper studs, incorporating an unthreaded section, would remain a challenge.

Not sure why;    …

Sorry, Andrew … badly worded on my part !!

All I meant was that my proposed “make-from” source of M3.5 would not provide a particularly simple way of making the studs.

MichaelG.

[Nigel Graham 2Participant @nigelgraham2]

Andrew –

That is exactly what I had described – and it’s an old technique!

 

Duncan –

No, don’t lock the stop-screw, but use one long enough to fit with its own lock-nut. Its should be left free because the action of cutting the second thread, especially with a die, will tighten the stud against the stop-screw, so you need be able to release that if necessary.

I am otherwise assuming removing the finished stud from the holder by the two-nuts method.

It also allows some length adjustment if necessary, including using the same holder for two or three different lengths.

[Nigel Graham 2]

On Nigel Graham 2 Said:

 

…should be left free because the action of cutting the second thread, especially with a die, will tighten the stud against the stop-screw, so you need be able to release that if necessary.

I am otherwise assuming removing the finished stud from the holder by the two-nuts method.

I just used a Mole wrench; a close look at the second end threads in the first picture above show some minor marks. But it doesn’t matter a jot as the thread will never be seen once the studs are loctited in place, and the method is fast. My threads were cut with a Coventry diehead. I don’t know if that causes more, or less, tightening than a circular die.

Andrew

[Nigel Graham 2Participant @nigelgraham2]

It may be fast but I’d rather not risk scarring the surface or worse the thread, even if it is hidden.

I do have a Coventry die-head which I have yet to use, and I am not sure if my collection of dies is actually for the head’s size anyway, but any sort of die will impart considerable torque to the stud. It does depend considerably on the material toughness, diameter and thread dimensions of course, but the tool will still try to push the stud round hard against the back-stop.

[Nigel Graham 2]

On Nigel Graham 2 Said:

It may be fast but I’d rather not risk scarring the surface or worse the thread, even if it is hidden.

I do have a Coventry die-head which I have yet to use, and I am not sure if my collection of dies is actually for the head’s size anyway, but any sort of die will impart considerable torque to the stud. It does depend considerably on the material toughness, diameter and thread dimensions of course, but the tool will still try to push the stud round hard against the back-stop.

Having had a quick look at some of my spare studs any thread damage caused by the Mole wrench is not immediately obvious. Clearly the diehead does tighten the stud against the brass backstop, but it doesn’t take much to release it. I’m lazy and when making dozens of studs I’m not going to mess around using the double nut method. Grin smiley

Andrew

[Nigel Graham 2Participant @nigelgraham2]

I take your point about convenience and speed! Anyway you can always protect the surface with some shim material, e.g. cut from an aluminium drink or fish tin.

[Author]

Posts