Where is the best thread making video featuring correct method?

having seen comments on folk getting the art of thread making wrong, where is the best video with correct technique for this ?

To include the nuts thread making as well.

I am having to make a 3/8 inch British Standard Fine thread. bolt.
I find the attached charts.

I also have something larger !

threading need

the angled tip is clear of thread, it fades (runs out) into its lower area. Its that end that gets screwed into the threaded hole.

I attach the body needing a thread,
ext dia of the thread is that of the item at 1.055 inch.
pitch is 0.05
valley of thread 1.00 inch dia

even at this fine thread there will be radiused tips and rules to obey.

amazing is thread making, superb engineering.
thread to make.3dm (118.7 KB)


1 Like

Hello- if I am looking at this right, this does not work out… if the depth of the valley is at .5 radius, then the thread radius there is ~.01 - if that is the same at the peaks, then the line tangent between them is not on 27.5 degrees, or the pitch cannot be .05.


1 Like

Hi Pascal,
If we let the theory decide the distance from valley across to valley on other side, what does that make it ?

What I do know for sure is the pitch is 0.05 by measuring the item very carefully, even photographing a vernier gauge teeth offered up to the thread , as well as taking a measurement of 10 peaks which came to 0.5 inch. thus 0.05.

visually I saw 1inch side to side at the end but it might be slightly different.

taking a few diameters it was usually 1.055

another item that screws into the same female thread had been 1.06

So what if the valleys is decided by the maths, what should it be ?

I do some maths, the d is .03201635

x2 and thats .0640327 and 1.055 less that is 0.9909673

as I measured 1.0 by eye, that shows I wasnt far out !

so valley to valley is 0.9909673

what is interesting is the table shows 10tpi for 1inch dia.

this item is not 1 inch but 1.055 and has 20tpi not 10.

same as the 3/8inch data.

The tables are for bolts perhaps, my item is not a bolt.

I need to see a vid showing correct best practice for creating this thread as per initial post, how to fade it out into the 45deg slope.

also how to do a bolt. and nut thread. the table has the slight differences to allow one to screw over the other.


1 Like

Hi Steve - if you remove the depth constraint then you can make it work out - I suppose it is something like this -

thread to make_PG.3dm (234.6 KB)
The thread is made from a RailRevolve, you can stack as many of these together as you need to make a thing that you can BooleanDifference from the nut or bolt.

1 Like

Hi Pascal,

I see the theory, very clever. I like that. :smiley:

How does one get the thread to feather out at either end ?

I always struggle on that as do many.

making a helix one has to bend the line inwards but there are good and bad examples on the correct bend inwards to make and I was looking for a video on a good one.



Hi Steve- you can make a taper but I have no idea what the correct geometry is for that. Extend the helix with an arc that dives slightly inward and then sweep the edges on the end of the final thread on that.


1 Like

there is a very famous german television show for children - and one of their movies shows very precisely how screws are made:

the thread is pressed into the rod.
this will also lead to more strength instead of cutting the thread.
…this means the final ending of the cut / helix - V-shape is given by this mechanical process.

link to the other already very elaborated topic about nuts…


It isn’t a BSF thread. It may be a BSC 20 thread (British Standard Cycle 20 TPI) or, more likely, a proprietary thread as 1-1/16 diameter is an unusual size. The profile is likely to be different from BSF. For example BSC uses a 60 degree angle.

I would try to examine it closely and see if it has radiused crests and valleys or not. If proprietary, it may simply be turned with a simple v-cutter and hence flat sided.

1 Like

Hi Pascal,

I study the thread with x10 magnify and see it takes an entire turn to disappear from being normal depth and diameter.

How best do I do that for the last ‘lap’ ring of the helix ?

Cycle thread seems a good thought. The item was made by BSA and they also made cycles.

its been coated with a fine preservative so analysing if its got any radii valleys is not possible as the fine red liquid has given it radiused valleys ! though the peaks are quite sharp but at 20tpi its tricky anyway.

another one in use and unscrewed a bit is shown here, quite sharp looking.



1 Like

Please anyone ?
thread to disappear into the solid by the time 360 degrees rotation is reached, how is that done ?


1 Like

Please ?

I have a fillet thing to finish and a thread, and 2 weeks behind just solving these two, getting knowhere… and surely someone knows how to sweep a thread inward during its next 360 deg rotation ?

In other words if one has a helix, how is the last rotation taken progressively inwards during the 360deg rotation of it.

I know someone out there must know this simple thing ?


1 Like




Project it on to an appropriately designed cone?

I didn’t realize you actually got this specific at one point before?

Should the video teach the audience the workflow of helix commands for various thread profiles?

Or should it get really specific on how to interpret the machinist handbook as well?

Machineshop 101, a ‘thread relief’ is created prior to the threads being cut?

That’s not what machine shop 101 teaches. Think of how the cuts would have to be repeated on a manual lathe.

Of course in Rhino you can modify helix’s and any line/curve to do basically anything. And you can program CNC machines with CAM etc, to attempt almost anything.

What’s the design intent? Are there optical illusions?

Here is the item I wish to make and have machined so as to fit into the threaded holes that exist for it.
Jeremy has shown here
link is:-

that the thread doesnt need to move inwards within one rotation of itself as the cone end takes care of the thread appearing, this ties in with comment on the lathe just makes a constant diameter thread.
Note the tapered end and that this appears to fizzle out on the slope.

lower pic is a quarter turn more CW as I held it to the upper pic.

Images mobile phone SE2020 so not as nice as my Canon 60mm macro !

thread is OD 1.055 inch, shaft is 0.981 inch dia pitch 20 tpi

I dont see any charts with this diameter, as this is a thread made into an item the charts never were aware of and is not a standard item but somewhat specialist. Its BSC cycle thread it seems, made by the same company that did cycles.

A video on using a chart and making a bolt as on the chart would be meaningful to folk.

Following Jeremys cone end. Here the cone isnt 45deg but if 45 deg is a known angle then follow 45 deg.


1 Like

Don’t know if this would help, but I made a thread-generator in Grasshopper. You can find it here:

Some pros and cons:
(1) It needs some minor updates, which I’m hoping to do soon.
(2) As input, it needs a thread axis (specified by two end-points) and a thread-profile.
So, once you have the thread profile (tapered or straight), it will generate solid objects that represent the thread. But that does mean you need to create the thread-profile yourself (as a curve).
(3) Right now, the “Tolerance” input says “mm” but is actually in whatever units your document is using. I need to fix this.
(4) There are standard tolerances for thread-fits, and you would need to account for those in the positioning of the thread-profile relative to the thread-axis, as well as the numerical input for the tolerance.
(5) On extreme cases, this GH script will fail.

If you find any bugs or have suggestions, let me know. I mostly created this GH script as a way to learn Grasshopper, and for creating custom threads. But in theory, there is no reason it couldn’t be used for creating standard threads.

–Anthony K. Yan

P.S. Based on how a tapered thread is cut on a lathe, I think it is impossible to make a geometrically-perfect tapered-thread using Sweep1 and/or Sweep2. So instead, the thread-generator uses Loft.

The reason Sweep1 and Sweep2 fail is not obvious. This is geometrically subtle (unless you are a mathematician and/or engineer). Here’s the short “explanation” that isn’t an explanation, unless you understand the problem already. On a lathe, the thread profile is always in the plane of the lathe axis. That’s because the thread-cutter is always in the plane of the lathe axis. The lathe-axis is unique and global.

However, Sweep1 and Sweep2 orient the thread (sweep shape) based on local geometry of the rail(s). Even if you use Sweep1 with “Roadlike” orientation, you will get sweep-shapes that are not in the plane of the lathe axis. “Roadlike” means the thread-profiles are in planes that are parallel to the lathe axis, but do not always contain the lathe axis.

Sorry if this explanation is obscure as all heck. If there is interest, just ask, and I will make some example files and images to show the problems with using Sweep1 and Sweep2.

Thread relief is usually cut before the threads are cut. Generally, the thread relief is a cone (either Boolean subtraction or intersection with a cone). I think most of the threads I see use a thread-relief of 45 degrees. But personally, I like to use a 60 degree cone (120 deg included angle) which is the same angle as in the thread-profile. I’m not an engineer, so not sure why one relief-angle would be better than another.

I wasn’t aware of the RailRevolve command. I’ll have to try it. I think it does the right thing for making straight threads. Will have to try it and/or think if it works on tapered threads.

1 Like

Why do you need to model the thread? The only information needed to machine the threads are the diameter, pitch and thread profile, and of course the overall dimensions and shape of the part. I doubt the machinist/engineer machining the threads would even look at the 3D model detail of the threads.


Does it matter from any design or mechanical perspective how this thread is started?

In reality, you have a few options, all of which wil resolve your problem from a practical perspective:

  • Define that the thread start has an n-degree chamfer cone, which is what your image seems to show. Modelled using the original cone boolean response.

  • Define a new higbee thread which can be produced either from a synchronised milling process, or a turned process (using a grooving insert on a CNC). Modelled in rhino using a large diameter cylinder at the end.

  • Request a manual deburr operation at the end, which would produce no cone, and you wouldn’t define in the model.

If it doesn’t matter, just leave a note on the drawing to leave this up to the CNC programmer. They will likely elect for the easiest option, seeing as it is barely over an inch, and nothing like a thick buttress or ACME thread.

If it were a couple of inches or more, and needed a “heavy duty” thread, it would then perhaps need more thought towards a proper mill-syncronised or grooved higbee.

1 Like

I actually started to make a video tutorial couple of weeks ago, but Rhino crashed at some point, so I will have to repeat the process. Most likely I will record and upload my tutorial on YouTube tomorrow.

1 Like