Threads are complicated subject, and pre-standard threads are a complete muddle! I don't see any alternative to measuring their pitch and form.

I highly recommend Tubal Cain's "Model Engineer's Handbook" because it contains most of the technical data needed in a small workshop in digestible form. Section 4 covers standard threads: Whitworth, BSC, BA, Unified, Holtzapffel and Metric. Near the end there's a table summarising lathe infeed data for Metric, UNF, Whitworth and BA, note that internal and external threads are slightly different, and that threads can be sharp or rounded.

The infeed is related to Pitch, so step one with a TPI thread is to calculate it. If the ELS is programmed in inches, the pitch is simply 1/TPI. Otherwise, convert inch pitch to metric 1/TPI*25.4

The infeed varies by system, exact values in the book, but for an external thread the infeed is about:

• 0.6 x pitch for rounded, or
• 0.8 x pitch for sharp

I rarely cut big threads to exact specification, and – so far – they have all been metric. So I infeed by about 0.6p or a little more. Most fasteners are a loose fit; more care would be needed if I was targetting a tightly specified thread.

My approach to medium threads is to lathe cut them to about 0.5p and then finish off with a die. The lathe does most of the work and guarantees a straight helix; the die finishes the thread to correct form.

Not my thing, but I've successfully made the equivalent of a thread-plate for a non-standard thread:

• Determine the pitch and diameter of the wanted thread, usually by measuring it
• Set the lathe up to cut that pitch, or as close as the banjo & available gears allow.  (With luck it will be 'good enough' for a short thread
• Cut a sharp thread into a length of silver steel of the correct diameter.
• File or mill the silver-steel lengthways to make a tap and harden
• Tap drill a larger diameter disc of silver steel, countersink one side slightly, drill or file 2, 3 or 4 lengthways notches and use the home made tap to thread it. The disc, (or strip of gauge plate if a thread-plate is wanted), need not be as thick as a commercial die. Then heat harden.

Don't expect the result to be long-lasting or tip-top, but the tap and die will both cut provided the lengthways notches provide a sharp edge with some relief.

Dave

Edited By SillyOldDuffer on 21/05/2023 09:40:26

Depth is very dependant on the tip of the tool used, unless you can accurately grind the very small tip radius for the given thread form then trying to put a number to the depth so the flanks end up in the right place is going to be near impossible.

Throw into the mix tool and workpiece deflection particularly with small threads, backlash, etc and things get even harder.

Edited By JasonB on 21/05/2023 10:02:36

Quality inserts are generally ground to a specfication so they cut a book thread with book infeed. As do the Johannesson / SKF / Dormer 3/4 circle chasers I prefer when screw cutting external threads. Which makes life easy if you re lucky enough to have them.

Hand ground tooling will almost certainly be too pointy so extra depth is needed for clearance. Which can either be added as an empirical, usually works, factor as Dave does or you can try being a bit more scientifically investigative about things as I do when using the zero-to-zero threading method. Functionally zero-to-zero is pretty much equivalent to using an ELS for threading as you stop at a depth pre-determined during set-up. If that initial depth turns out to be too shallow I make successive slightly deeper passes until sufficient clearance is obtained for parts to screw together. Just like CNC and and ELS the zero-to-zero method lets me add feed in accurately determined increments so that once success has been achieved I can adjust the set-up parameters to compensate so all subsequent threads are right. In an important application I use increments to something close to the specified thread tolerances so the difference between the last doesn't fit and first does fit threads will be small enough that the does fit one will be within tolerance.

Alternatively you could cut flat bottomed threads where the root is flat rather than curved. The flat width being sufficient to encompass the correctly shaped thread from end. Using calculation and drawing board or CAD techniques the difference in depth between a sharp point and a flat of the correct width can be determined. As can the infeed depth needed. Creating the tool is a simple(?) matter of grinding a sharp point of the correct angles then flattening the end by the desired amount. Something that is easy to keep track of by measurement.

Clive

As Jason says even the flat end tool is by no means easy to grind in smaller sizes.

Objectively you are in micrometer adjustable jig territory. The great advantage is that measuring how much you have taken off the sharp pointed tool you start out from is relatively easy and, if the flanks were jig ground at the correct angle, the width must be right. So even if things don't work you know exactly what you have giving a reasonable change of figuring out what to do next.

Being able to reliably get "the same" wrong result is always a better start than wondering "what went wrong this time?".

But it all starts on the drawing board or CAD screen. Home Shop Guy hasn't a hope of measuring this stuff so all he can do is carefully work out what is needed then establish reliable techniques to do it.

Had a few "D'oh twot!" moments after carefully plotting out what I'd done on the CAD to prove that it was right only to discover the connection between the effects of what was done and correct was somewhat nebulous.

Clive

Dell

The only way I've ever been able to get this sort of thing clear in my mind is to draw things out accurately as overlaid diagrams so as to see exactly what the dimensions need to be if things are to fit.

CAD makes this so much easier. And clearer if you use coloured lines. A decent CAD drawing will also easily give you the amount you need to take off the tip and find the length of tool plus flat tip that you will actually measure.

Clive

If it is not a fine thread and you can determine the width of the flat you need, a quick and not all that nasty technique is by using a diamond slip together with a reasonably good several times magnifier. Compare the flat as you stone with the small graduations on a decent quality steel rule, or with the threads on a small screw of known pitch. It is possible to judge down to a flat of about 0.1 mm. It is then also feasible to round to a passable radius. You can combine this method with some of the measurement techniques suggested above.

.

First and best [for measuring male threads] is probably an optical projector [Shadowgraph or similar] … but note that for accuracy ‘telecentric’ lenses are required, so a home-brew version is tricky.

After that [for the small sizes that particularly interest you] a digital photo with a good macro lens, or a microscope, is a decent place to start … plenty of software available to measure images to ‘pixel-dimensions’ and to calibrate to real-world dimensions.

[ Again, a telecentric lens is most useful, but ‘needs must, when the Devil drives’ ]

If you only have the female thread available, then take an impression of it … using wax, resin, or a low-melting-point alloy.

Have a think about those, and shout if you need anything more.

MichaelG.

.

I will find a demonstration of telecentric lens optics later … need to put the shopping away now !

Edited By Michael Gilligan on 21/05/2023 17:09:41

[ Shadowgraph ]

Classic versions, and modern alternatives: **LINK**

https://www.bowersgroup.co.uk/brand/baty

MichaelG.

Perhaps one of those cheap USB microscopes and you could have a close look at the thread angles on the computer screen.

.

Dell

Forgive me, please, if I have mis-understood, and therefore exaggerated, your requirements.

I thought you were seeking a way to accurately measure the pitch and profile of an historically important screw thread which had probably been produced with a unique set of tools.

Checking the pitch of a threaded hole, using peg wood, is fine … but it will not properly identify the profile.

At the risk of stating the obvious:

One cannot grind a single-point threading tool to the correct shape until one knows the required shape.

MichaelG.

.

P.S. __ How obsessive you need to be about this would obviously depend upon the importance of the clock and whether you are doing restoration or renovation.

Three wire thread measure.

You can check the flanks of the thread with three wires and a micromiter. I use the shanks of three drills as they are hard and ground usually. You can then check the part you are making against the original or new screw. If using a worn part use a piece which has least or no wear.

David