Measuring thread depth without wires - your views, please

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

HennieL

Well-Known Member
Joined
Apr 17, 2020
Messages
118
Reaction score
84
Location
South Africa
Hello everyone,

Here in South Africa where I live it is very difficult, and extremely expensive, to obtain the correct sizes and accuracy wires to use in the 3-wire method for determining thread pitch diameters, (and thus confirming that your single point threading is cut to the correct depth). So, until recently, I have been doing what most machinists do - screwing a nut onto the threaded bolt, and calling it OK if the nut goes on easily, but tightly...o_O

Some time ago I purchased a new micrometer that has a 0.001mm resolution "sight unseen" through the internet, and although I'm not very happy with the "feel" quality, it is giving me consistently the same (rounded) results as my trusty old Mitotoyo 0.01mm resolution micrometer - hence it is at least accurate to 2 decimals of a mm. Anyway, to make a long story short, this micrometer came boxed with a set of exchangeable "anvils", one set being 60º wedges, another a set of 60º cones, and a third set are wide, flat disks. Last weekend I opening the box to remove the micrometer in order to take some readings on a rod that I was turning down to make a special bolt, and the 60º wedge shaped anvils caught my eye just as I was thinking about the critically sized threading that I had to do... light-bulb moment... Why not use a flat disk anvil on the one jaw, and either the wedge or the cone in the other jaw of the micrometer - then I can calculate the thread depth by measuring the thickness of the (outside) major diameter of the "bolt" normally, and the diameter of the "bolt" with the wedge fitting into the thread on the one side, and subtracting the one measurement from the other should give me the thread depth...

Bit difficult to explain - let me show you some photos to make this clear...

The micrometer fitted with the flat disk and wedge anvils into it's two jaws:
Micrometer.jpg


This is a close up of the 60º cone and wedge anvils:
Anvils_1.jpg


And here is a photo showing the flat disk straddling the outside of the threads on the one side of the bolt, and the 60º wedge fitting in between two threads, measuring to the minor diameter of the bolt (M16x2).
Thread_1.jpg


So, is this an accurate way to measure screw threads???

I wanted to confirm if this idea is actually worth something, or just a "flash in the pan", so I raided my supply of bolts and machine screws, and measured these bolts varying from 20mm to 4mm diameter.

Bolts.jpg


Keep in mind that all the bolts were Metric, and all were bought from shops (and one could thus hope that they were manufactured to prescribed tolerances).

Results are rather encouraging, I think - I measured 38 bolts, and found that three of them had major diameters outside of the normal 6g tolerance, and these were excluded from the final assessment. As can be seen in the spreadsheet summary below, the average measured thread depth of all the "in spec" bolts came to 0.72mm, whilst the calculated thread depth was 0.71mm when using the formula: thread depth = dmax - d1max (also known as 5H/8), and 0.80mm when using the formula: thread depth = 0.6134P (which is the most used formula). For interest sake, I also looked up the recommended total thread depth for two carbide inserts manufactured by Sandvic and Mitsubishi. Lastly, I calculated the % deviation of each individual measurement, expressed as a % deviation from the two calculated formulae, and highlighted the calculated depth closest to the actual measured depth.

Data.jpg


I do accept that the wedge has a very narrow blunt edge - perhaps 0.25mm, and that there would be a smallest limiting size beyond which one would not get accurate readings - the above figures show that the % deviation increased substantially with the 4mm diameter machine screw readings, and this probably exceeds the minimum readable size. Also, there is no allowance for the changes in the helix angle resulting from the different pitch sizes, but I suspect this will be negligibly small.

I really would value any feedback on this: - Are any of my assumptions wrong? What am I missing? Why are the larger bolts closer to the 0.6134P values, and the smaller ones closer to the 5H/8 values? Why did I do this, do I need to see my shrink🙃
 
I use rubber bands to hold the wires in place. I only use this when a test thread for internal threading ( lathe chucks backs).

Most machine shops have board with every know thread and just feel using the nut. If not on board they ask for nut to fit the thread.

Update 6/13/2020
Kvom have a good point, use what you are thread too the may be a different class fit.
 
Last edited:
Thanks for the replies so far, but I'm afraid you're missing the point a bit :) (with respect...)

When the professionals (and many of us amateurs) have a high-tolerance threading requirement, the standard "go to" method of confirming if you are within tolerance is to measure the pitch diameter using the 3-wire system - but as stated in my original post, they are very scarse and VERY expensive where I live. Also, in my experience, one could calculate the depth of cut to 10 decimal places, and then still need to make two or three "cleaning cuts" to achieve the correct thread depth due to tolerances and flex in our machines (at least, in mine... my threads always come out shallower than what I calculated).

Anyway, if you really want/need to know if your threads are within tolerance you either use the 3-wire method, Go-No Go gauges, or a special thread micrometer. The only difference between the method described above and the professional thread micrometer is that I don't have a V-slot anvil for my micrometer, and substitute this by using a wide flat disk anvil that straddles a few threads. Truth be told, I did not even know how a thread micrometer looked until after I thought about using the 60º taper wedge and flat disk, and completed the tests shown above - only Googling for "thread micrometer" early this morning, and discovered that "my" method is quite similar.

Still anxiously awaiting some in-depth feedback / comments to the following:
Are any of my assumptions wrong? What am I missing? Why are the larger bolts closer to the 0.6134P values, and the smaller ones closer to the 5H/8 values?
 
Here a little trick for thread.
If making a thread where both ID & OD use same thread per inch as lead screw you do need the thread dial.

If lead screw is 8 TPI you use and thread 2X, 3X 4X, or basically any even multiple of lead screw that is finer.
Like 16tpi, 24tpi 32 tpi

If lathe has 12 tpi the same goes 24tpi, 36tpi or 48tpi

This works for your own project when you pick the thread pitch.

Dave

Thanks for the replies so far, but I'm afraid you're missing the point a bit :) (with respect...)

When the professionals (and many of us amateurs) have a high-tolerance threading requirement, the standard "go to" method of confirming if you are within tolerance is to measure the pitch diameter using the 3-wire system - but as stated in my original post, they are very scarse and VERY expensive where I live. Also, in my experience, one could calculate the depth of cut to 10 decimal places, and then still need to make two or three "cleaning cuts" to achieve the correct thread depth due to tolerances and flex in our machines (at least, in mine... my threads always come out shallower than what I calculated).

Anyway, if you really want/need to know if your threads are within tolerance you either use the 3-wire method, Go-No Go gauges, or a special thread micrometer. The only difference between the method described above and the professional thread micrometer is that I don't have a V-slot anvil for my micrometer, and substitute this by using a wide flat disk anvil that straddles a few threads. Truth be told, I did not even know how a thread micrometer looked until after I thought about using the 60º taper wedge and flat disk, and completed the tests shown above - only Googling for "thread micrometer" early this morning, and discovered that "my" method is quite similar.

Still anxiously awaiting some in-depth feedback / comments to the following:
 
I have 3 sets of decimal drills available to me (at different locations) from 0.3 to 1.6 in 0.05mm steps and 1 to 6mm in 0.1mm steps - so I choose the nearest appropriate shank size - do the calculations and off you go.
You don't even need three you can use say two 0.8's and a 0.75 (illustrated below) and calculate accordingly (or draw it up in a CAD program to save you the trig.) - measure the actual shank diameters at the point you intend to use as they are slightly tapered with your typical twist drill shank being 0.05 to 0.1mm smaller than the cutting size -diminishing with size.
I have two Microbox sets 0.3 to 1.6mm diameter in 0.05 steps which is small enough for most thread sizes.
I normally have two sets over my lathe so 2+1 is my go to method and CAD 'cause I'm lazy.
You obviously use a wire diameter that is going to register close to the effective thread diameter - but anywhere close is good enough if you do the calculations / CAD vis:- M8x1.25 nominal for illustrative purposes:-
1592131323786.png

Regards, Ken
 
Last edited:
With respect, how often do model engines require threads more precise than can be obtained by simpler methods than wires? The micrometer setup you have looks like a good solution for checking once you get close while single-pointing.

FWIW, McMaster has a set of 48 wires for $36.30. I'd be willing to bet that a member here would be willing to order a set and ship them to you at cost.

https://www.mcmaster.com/thread-measuring-wire
 
Thanks for the replies so far, but I'm afraid you're missing the point a bit :) (with respect...)

Still anxiously awaiting some in-depth feedback / comments to the following:
I'm following you, it seems quite logical me, but I'm only a novice.
Cheers
 
Hennie,
Yarwellnofine (that makes sense if you are South African) - I get the question - yes the method you are suggesting works but the crest radius / flat (and that varies greatly) is going to interfere with your measurements.

If you are single pointing the crest is going to be sharp - I often screw cut to the sharp point at full diameter as my "gauge" point.

Using your micrometer set up is going to work just fine if you work to a truncated (flat) tip profile like ±7.8 diameter in the case of an M8.

Problem is how do you calibrate the nominal starting point ? Sure if you know the precise tip radius of your anvil you can work it out and calibrate against a known thickness (slip block ?) or do you have a calibrated "V" - again this "V" can be calibrated using a wire.

A simple calibration piece would be a thread with a precisely finished O.D. and therefore a truncated thread (flat) profile - you could use any thread that you have measured - via the three wire method - and documented. However you would still have to adjust for the truncation of the thread you are turning versus the thread you calibrated against.

If you work in a production environment and can get your hands on a selection of No-Go thread plugs these would make excellent calibration pieces.

I think your micrometer is just fine for production methods but for one offs it is probably more trouble to calibrate than it is to use the three wire method.

And all the errors of angle and symmetry etc. etc. are still there to haunt you regardless of method although the 3 wire method kind-of evens these errors out a bit.

In practice I normally screw cut to the theoretical sharp corner and test for fit against a nut or bolt / mating part as required.

That micrometer is a great piece of kit - did you make the wedge / point inserts ? Or did it come like that ?

Regards - Ken

P.S. All those commercial bolts were rolled threads and therefore the resultant finished effective & major diameters were all influenced by the diameter of the rod prior to rolling.
An undersized blank will result in excessive crest truncation and undersized major diameter but the effective diameter should still be O.K..
An on size blank should produce perfect results.
An oversize blank will look perfect but be oversize on effective and major diameters.
Trust me - years of thread rolling experience on millions of parts - I could go on.......
 
Last edited:
Being a novice is just having how to. Back I started there was no internet. You had only books and just did not say the finer details.

When first started did not of know of thread wires, I use used a nut.
Later I did try use drill bits for thread wires but it was hard to use and was pain.
Once try using rubber bands witch made a lot better using thread wires.
I was still highschool working RC planes.

Dave

I'm following you, it seems quite logical me, but I'm only a novice.
Cheers
 
Thanks guys - some good points, and you made me think some more...

Why are the larger bolts closer to the 0.6134P values, and the smaller ones closer to the 5H/8 values?

I had a closer look at the wedge seating in the threads, and I think I discovered my own answer :cool:🤓

After checking with a strong magnifying glass, it's clear that the flat on the bottom of the wedge "bottomed out" against the side of the thread on the smaller pitched sizes (1.0mm and smaller), and not on the bottom of the thread - hence they did not measure the full thread depth as I originally thought. The larger thread sizes allowed the flat bottom of the wedge to touch the bottom of the thread, and thus agreed with the correct method of calculating the thread depth (0.6134P)... Oh well, another "light bulb moment" gone to waste :oops:😢

No matter, this was yet another step in my continuous learning process...

If lead screw is 8 TPI you use and thread 2X, 3X 4X, or basically any even multiple of lead screw that is finer.
Like 16tpi, 24tpi 32 tpi

If lathe has 12 tpi the same goes 24tpi, 36tpi or 48tpi

Thanks Dave, but having been metricated for more than 50 years (and living in a republic 😁 ) I try to avoid anything imperial if I can - so it's strictly metric threads for me... and my leadscrew is 3mm, for what it's worth - but I do hear what you're saying.

Yarwellnofine (that makes sense if you are South African)

Shapshoot my Bra 😀 (and that's another indigenous South African saying... apologies for this, but we're nearly as bad as the Auzzies)

Problem is how do you calibrate the nominal starting point ? Sure if you know the precise tip radius of your anvil you can work it out and calibrate against a known thickness (slip block ?) or do you have a calibrated "V" - again this "V" can be calibrated using a wire.

Ken, thanks for your input - you made me re-consider, and do some more research. I think I will continue to use the "wedge micrometer" on threads of 1.5mm pitch and larger, and will use the "one-wire" method to check if I'm in the ball park on the smaller diameters.

With respect, how often do model engines require threads more precise than can be obtained by simpler methods than wires? The micrometer setup you have looks like a good solution for checking once you get close while single-pointing.

Thanks for the offer kvom, I might just do that. Problem is, though, with our "junk status" currency that's at least 40 Bic Mac burgers, and I would still like to figure out another (cheaper) way, if possible. I do have a few sets of drills, though, and I'm going to try the "one wire" method using measured drills for a while and see how that compares to my micrometer (and my not-so-trusted nuts LOL).

As for your question regarding why I need the precision - well, that's just how I am, I suppose... seriously, though, I have been doing other "hobby" work with the lathe that do require accurate threading from time to time for some years now - but truth be told, being new to this hobby and to this Forum, I have not even made a single engine yet, so I cannot argue with you on this one... but wait, it's going to happen soon 🤠😁
 
I think this is a good way to measure threads if you use the actual o/d of your work piece, and grind a small flat on the wedge so it will contact the effective diameter and not the radius in the root of the thread. I will be trying this method next time i cut a thread.
 
I think this is a good way to measure threads if you use the actual o/d of your work piece, and grind a small flat on the wedge so it will contact the effective diameter and not the radius in the root of the thread. I will be trying this method next time i cut a thread.
Thanks Peter.

After giving this some more thought, I agree, IF you can make a whole set of wedges, one for each pitch - with the flat being approximately equal to the pitch diameter (same as the "best wire" diameter used when measuring with three wires). I'm more inclined to grinding a very small radius on the end to the same diameter as the insert that I use (0.3mm, if I remember correctly), and still try to measure the root. Having made custom knives for many years I have the tools (and grinding belts and papers down to 3000 grit), so putting a small accurate radius onto the wedge should not be too difficult. Hmmm, perhaps I should make some new anvils instead, and try both methods...
 
HennieL: I recently ran into this same issue with wanting to know that I made a correct thread without guessing that I followed the procedure of feeding the compound slide the correct amount, as every description I read, I felt confusing.
If I start with a sharp 60 degree cutting bit, and feed the compound that is set to 30 degrees off axis (~29.5), and I start with the round OD right at the described hardware diameter, I should feed the compound in the exact distance of the pitch of thread, as this is one side of the equilateral triangle. Ignoring the crest and trough radii. This should be a perfect sized thread (I understand a nut with the trough radii per the specification, will not fit on this cut thread). I've not seen a logical treatment of this. As it appears that any of the internet sites I find, talk about feeding in the cosine of 30 degrees times the pitch of the thread, which makes sense if you feeding in the cross slide, not the compound. So I'm still not understanding this single point process, and in the end, cut until a nut fits on my threads.
My solution was to buy a screw thread micrometer, in the 0-1inch size. This came with 5 sets of anvils.


pitchAnvil Pair #1Anvil Pair#2Anvil Pair #3Anvil Pair #4Anvil Pair #5
Threads Per Inch64-4844-2824-14 13-9 8-5
Metric 0.4 - 0.5mm0.6 - 0.9mm1.0 - 1.75mm 2.0 - 3.0mm 3.50 - 5.0mm

You have to zero the micrometer when you install the anvil for the size of interest. You zero the thimble reading, then push the bottom anvil into the mate, and lock it in with it's lever, as it is on a sliding foot.
But if you have problems getting the set of wires, you probably can't get this micrometer either.
 
When I was machining (on a lathe) in one of my jobs we used thread parallels, these are very similar to the wedge idea of yours, but you use a single wedge on top and a double wedge on the bottom these are very good to use and accurate, they are in sets. You put the pair together you are going to use measure them add this to the effective dia. of the thread you are going to cut and that is your size. if you google thread parallels you will find them. There is some on ebay but only BA ones.
 
I still think your single wedge is good and i an going to try it, but i will put a clearance flat on the point not a radius because you need to measure the effective diameter, if it sits on the radius you will get a false reading.
 
Last edited:
Instead of all the alternatives that require a lot of calculations and/or purchasing more parts, why not make the 'bottom' anvil for a thread mic. you have a lathe and a mill so it's probably worth a try. Here's a picture of what they look like and given what you say prices are, you could spend quite a bit of time making some of these up, hardened drill rod or high carbon steel would work. They would probably even work unhardened as they have to be zeroed each time they are put in the mic. getting the vee anvil centered would be important.

pic.jpg
 
Instead of all the alternatives that require a lot of calculations and/or purchasing more parts, why not make the 'bottom' anvil for a thread mic. you have a lathe and a mill so it's probably worth a try.

Thanks for the replies everyone - lots of good suggestions.

L98fiero, I plan to make some more "top" wedges with different edge radii to accommodate the various pitches - and I will probably make them from M45 tool steel, so I might as well harden them too. Making the "V" anvils to the required accuracy would be much more difficult, I think - unless one makes some properly shaped "inverted cutters" first and then cut the "V"s... Probably more work than it's worth. Will give it some thought, though...

By the way,your pic does not show...
 
By the way,your pic does not show...
There are no dimensions where I got the picture(Google search) but the anvils are for a range of threads so knowing the range one could work out the pitch of the crests on the anvil. Just a guess without checking my set would suggest the pitch to be for the finest pitch of the range and they could be made by tilting a rotary table 30° and doing 2 passes with a hand ground fly cutter.
 
Hi Guys,

An easy way to make some anvils would be to drill a suitably sized hole in some round material and thread it, and then mill just over half off the thread. This will leave you with as many thread forms as you would need and the correct radius for that thread size. When done they would look like those in the picture above.
 

Latest posts

Back
Top