View Full Version : Threading on the lathe

05-04-2006, 11:20 PM
Posted this on the "Third Hand" but on ly got one reply in two days.
I am a beginner so please explain in simple terms, Thanks! When looking in the Machinists handbook, I see for a 1/2" - 13 class 2a external thread the major dia has a spread from max 0.4985" to min 0.4876" and the pitch dia max 0.4485" to min 0.4435" 1st, What is does the 0.0015" in the Allowance column mean 2nd, What is the 0.4822" in the min (small d) column. 3nd, How do I decide where within these spreads to shoot for when cutting the thread. Will external threads work properly with a corresponing same class intenal thread if both have finished dimensions that are anywhere within their respective tolerance spreads. Thanks, berol

05-05-2006, 12:09 AM
As I told you before, don't drive yourself silly with this. Unless you are goind to buy an expensive set of thread mics or adapters and do a lot of calculations, use thread wires and the chart that comes with them. You won't go wrong and if the hole is tapped the male will work if cut to the wires. Unless your cutting ultraprecison threads for the Hubble telescope, this will get you through fine. Hey, its worked on parts for nuclear reactors on Navy subs so it should be good enough for what your doing.

05-05-2006, 11:02 AM
Machinery's Handbook as well as many texts go into great detail about classes and fits of threads. For a beginner or HSM, it can be overwhelming and seem very technical. From a practical standpoint such as a one off project, all you really need to do is thread to fit a standard nut using the nut as the guage. For an internal thread, use a threaded bolt or screw as a guage. As previously said, unless you are making parts for the Hubble telescope, this will probably suffice for 99% of the work you will do. I have been doing it for years this way.

Perk in Cincinnati

kap pullen
05-05-2006, 05:06 PM
I used to do lots of bearing retainer nut threads on shafts.
Got tired of losing the wires in the pan.

I sawed two pieces out of a purchased bearing nut to about 3/4 inch long.
I measured over the nut sections held on a properly threaded part, marked the dimension on the nut pieces, and used them as a gauge from then on.

You could saw a 1/2 - 13 nut in half, measure it held over a known good thread, and use it as a gauge from now on too.

Even NASA frequently makes their nuts to fit existing parts since someone mentioned HST.

I did have to make some thread gauges held to a -+.0001 tolerance.
A couple parts went in the pan to get those right.

Kap Pullen

05-05-2006, 11:12 PM
Berol, I'll give you my spin on this nut thing. Imagine a nut and thread that have 100% contact, sharp v points and no clearance between the parts. They'd pretty much weld together as soon as they met. Not a very practical mechanical arrangement. On Class 4 threads, the nut will start a thread or two by hand, then will need mechanical advantage to advance any further. Class 4 tends to be pretty permanent when it gets where it's going as well. Class 3 and the nut will mostly start by hand, but somewhere near the end will need a wrench. This class stays together pretty good too, but can be coaxed apart with normal tools. Class 2 will spin up to the shoulder by hand. Class 2 is hardware store variety fit. Class 2 is about 75 to 80% contact if memory serves.

Now back to square one. Ya turn your stock to .500" and that's too big. Them pointy points are cutting up yur fingers, too. Ya turn your stock to .495 and now you've got 2 1/2 thou breathing room per side. That ain't much but it's better than nothing. Turn it to .490. Roomier still. With each decrease in diameter, the V is getting a little flatter on the top. Can you see it? The flanks need to stay perty close to keep the nut from wobbling.

All them charts will make YOU nuts. Turn your OD to .495 so you know it clears. Note your starting points with the compound and crosslide. Cut the threads until the bluing just disappears and you feel a burr raised on the point. Try the nut, it should start but won't go. Knock the burr off with a couple passes of a file WITH A HANDLE ON IT. Now your nut should go. That is, if you got 48 other things right to start with. Corrections can be made with light passes with the file, the threading tool, or advancing the crosslide without moving the compound if the flanks of the threads are still too tight. Make some test pieces, have some fun! Single pt. threading really is one of the most enjoyable chores you can do on a lathe.

Now, my friend, do yourself a favor. Make a notebook with a page for each thread you ever think you'll cut. This page will be 13 tpi. Write down the actual major diameter it ended up, the angular depth you went in with the compound, and what class fit you think you got. And any other notes that might come in handy, like if it was a commercial or hand ground threading tool. In a few years you'll have a pretty good track record of what it takes to cut 13 tpi. That angular depth number will be pretty similiar each time. This comes in REAL handy when you want to do some critical work. Let's say your average angular depth is 48 thou for 13 tpi. I'm making this number up, don't take it verbatim. You'll know in the future you can run up to 44 or 45 a.d. with impunity. A bunch of jobs get ruined by worrying too early, then getting frustrated at the end when you really needed the patience.

If you like math, study them charts. If you like making stuff, try it my way. All the best to ya. Gizmo

05-05-2006, 11:56 PM
Hey Gizmo2,

That is the best practical explanation of thread fit that I have heard. That is exactly the way that most of the old time machinists did it and after doing a lot of single point threading you get a "Knack" for the feel of a proper fit. It is the same kind of thing as what speed and feed you use for an end mill. You start with the chart and adjust up or down to get the desired color of chip. This kind of thing is the hardest thing to explain to young folks interested in learning the trade. Not everyone can just read the books and really understand all the information they have just gotten.


Jim (KB4IVH)