what od for a external thread

[Don Huseman]

As an example what should the od be for a 1/4 20 thread be to make the top of the thread flat . I know that you should round of the tip of the threading point tool, but what do you do for the sharp v crest of the thread.
Don

 

[tattoomike68]

On smaller threads I cut the diameter .005" under and on large threads like 1.250" or bigger I cut them .010" under nominal or even more.

 

[rake60]

Using your example of a 1/4-20 thread the book calls for an O.D. of .245" +./0 -.003
With the raised burr it will be a sharp edged thread, but cut it will fit a 1/4-20 size piece.

I turn the O.D. to .010" under the nominal size.
The crest is not as sharp and the thread integrity is not lost.
Please understand I have no engineering experience to back up my personal
methods! Just the experience from past years. It's just what works for ME.

Rick

 

[tattoomike68]

After cutting threads I always take a file to them and knock off the burrs too. that knocks the sharp edge off.

 

[mklotz]

The height of the flat is H/8 where

H = sqrt(3)/(2*N)
N = thread pitch (tpi)

so the diametrical undercut from nominal size is

U = sqrt(3)/(8*N) = 0.216/N

which, for N=20, is 0.011" (good call, Rick).

A long time ago, I worked out the formulae for most of the pertinent measurements on a sixty degree thread form.

 

[Swede]

Very impressive! ;D I tend to just sneak a peek in one of my books!

One thing to keep in mind with threads... hobbyists tend to work to precisions that are often unnecessary, and worse, often make final assembly more difficult than it needs to be. I am guilty of one thing in particular - If I am connecting two pieces with, say, several 6-32 screws in a pattern, I'll lay them out normally, drill and tap the one piece, then with the other (requiring unthreaded clearance holes), I'll mike the connectors, and then grab the next larger wire-sized drill in my box. This tends to create problems, as the clearance between the screws and the through-holes can make alignment and final assembly difficult. I then have to open them up further to successfully put it together.

My point is that it is a known fact that the screw body itself will usually fail long before the threaded interface, so long as you have an adequate depth. I like to make life easier in this case by doing as Mike mentions, mentally taking off a certain amount for the OD blank, and for a 1/4" thread I'd go right away to 0.245" or less. Even one of 0.240" will be plenty strong, although a bit sloppy going on. It also makes using a die easier, if that's your plan.

But if the thread is for something like a needle valve, or some other important operation, then I will tend to take a bit more care and time on it.

 

[mklotz]

Swede,

I put the equations up there because they answered the OP's question. However, I have to agree with you about hobbyists going overboard on accuracy and "following the rules".

I seldom use the calcs myself - just file the sharp ends down until 'it looks right'.

 

[tattoomike68]

Tight threads will gaul, cross thread and generaly cause problems. A few thou of slop is a good thing 95% of the time.

One time I had a 6-3/16" 12 TPI ring gage get stuck on a realy big hydraulic cylinder tube (500 pounder)and had to machine the gage off and later build a new one.

Some slop would have been nice on that job.. ;D

 

[Philjoe5]

I'm fairly new at cutting external threads but I recall a comment in J. Harvey's "Machine Shop Trade Secrets" that recommends reducing the nominal diameter by about 1%, so a 1/4-20 thread would use a 0.250" - 1% x 0.250" = 0.247". I've gotten good results using this rule and since I can usually do this calculation in my head it's something I use all the time.

Cheers,
Phil

 

[rake60]

At work and at home I cut 60 degree threads to thread triangle sizes.

Using thread triangles mics the thread at it's pitch diameter.

If the micrometer reading is correct the proper depth is there and the
thread will fit any standard corresponding thread.

If reaching that reading results in sharp crests the O.D. was too big and
needs to be reduced to eliminate those peaks.
Is that over simplified? I don't think so...

They work very well even for very small diameter parts.

If you buy a set of triangles they come will all the literature to determine
the size the thread should mic over the triangles.

There is another method. Thread wires. Instead of two triangles that can be
attached to your mic with little rubber bands, it involves three wires that have to
be held in place to mic over. That's fine for an 8 pitch, but I wouldn't want to
try it on a 6-32! :-\

Rick

 

[mklotz]

Rick,

Not to argue with you but I was told by an old-time machinist that thread triangles can produce a (generally slight) error if the thread angle (60 degrees) is off by a bit.

As an honest-to-god-does-it-for-a-living machinist, can you comment on the truth of that? Note that this is for information purposes only. I'm sure that, for anything we're likely to do in our home shops, the consideration is totally academic and thread triangles are perfectly adequate (and a helluva lot easier to use than wires).

Some hints for newbies who want to try using thread wires...

Since you have to manipulate three wires and the micrometer, at first it appears that you need three hands.

For larger pitch threads, I've had success 'gluing' the wires to the thread with heavy grease before making the measurement. For the finer threads, I stick the wires into some kids modeling clay at the right spacing and hold the clay with one hand and the mike with the other.

The wires are short and of rather small diameter. Inevitably, you will drop one into the swarf and spend the next hour trying to find it. Before making the measurement, lay a clean piece of tin foil under the part so a dropped wire is easy to find. Another trick I've read, though never tried, is to paint the end of the wires with some bright fluorescent paint to make them more visible if dropped.

The 3WIRE program on my page will do all the associated calculations for you although most wire sets come with a cheat sheet that lays everything out for you.

 

[rake60]

If the 60 degree angle is not correct the you will end up cutting past the
standard double depth to get the reading over the triangles.
The thread will fit because the measurement is still at the pitch diameter.

The 3 wire method is easier to mic once the wires are in the thread.
For very small threads that alone is a enough to put me off.

The only time I personally ever use wires is when I'm cutting round thread forms.
Thrangles would bottom out of the radius in the bottom of a round thread giving a
false measurment in that application. The smallest thread I've ever used them on
was a 10 pitch API

 

[pmercer]

Once I've screwcut something, to get rid of the sharp outer edge, I speed up the rpm on the lathe and use a bit of 1200grit paper. This gives a nice smooth finish. I sometimes do this on die cut threads as well if they are sharp.