Lathe threading math

All discussion about lathes including but not limited to: South Bend, Hardinge, Logan, Monarch, Clausing and other HSM lathes, including imports

Moderators: GlennW, Harold_V

User avatar
Harold_V
Posts: 20248
Joined: Fri Dec 20, 2002 11:02 pm
Location: Onalaska, WA USA

Re: Lathe threadiing math

Post by Harold_V »

Well, I don't think there is a simple answer other than measuring.

For one thing, it is perfectly obvious to me that infeed alone is totally meaningless.

1) unless you know exactly what the OD is, infeed has NO REFERENCE, and so indicates no size.

2) Infeed does not account for spring, nor for screw error, nor for errors due to less than perfect form, off-center height etc. obviously the latter shouldn't happen, but spring means you don't know your infeed is the same as the actual cut.

Essentially, infeed is "open loop".

If everyone concerned in this matter understood what you just said, we wouldn't be having this conversation, Jerry.

Per wires....

As I understand the thread wires, each TPI / Pitch must have a specific wire size, or it will not contact at the pitch line (which would actually be at an angle to a diameter across the wire, btw, due to the V angle).

That is correct, but because the micrometer "sees" the wire only at the centerline of the part, the mike has no idea that the wires are not parallel. I can only assume that the slight variation in the distance across the thread due to the helix angle is considered a non-issue. I've never addressed that issue. Not sure it's ever come up!

A small set of "fits all" wires is probably only minimally better than the triangles, as it too will assume good form if it contacts off the P-D.

I would certainly agree with that. That's why the G-men don't like them.

Speaking of the triangles vs wires: It seems that the wires might also lie, as a thread might be perfect at the pitch diameter, and wrong elsewhere (for instance a "bulgy" V wouldn't interfere elsewhere).

The wires don't lie, but they must be sized as you suggest, therefore wire size is critical. Threads are measured at the pitch line for proper fit, with the assumption that thread form is not out of tolerance. I recall that the angle has only ±15 minutes of angle. On small threads that translates into only tenths. So long as you measure at the pitch line, the thread is considered good, assuming the form is as well.

It seems that th thread must be inspected both for pitch diameter and for basic form in order to be known correct within tolerance.

Exactly.

A gage should pick up most such problems, but I think even it needs a form inspection, unless it has really comprehensive go and no-go criteria. Might take a couple gages, but just one look on an optical comparator to verify form (and maybe even P-D.)

Thread gages are made with that in mind, and they are used in pairs as you suggest. The go gage checks the major, minor and pitch diameter for maximum size, and in an indirect way, also checks the form. The truncated no go checks the minimum pitch diameter at the pitch line down to the minor diameter.. If the thread form is off, by the time the go gage fits, the no go usually does, too. They're pretty cleverly designed, the major problem being that they don't provide dimensions so you know where the error is, assuming the thread is not correct.. but a little interpolation usually provides the answer.

Harold
Wise people talk because they have something to say. Fools talk because they have to say something.
Rich_Carlstedt
Posts: 1757
Joined: Sat Dec 21, 2002 12:16 am
Location: Green Bay Wisconsin USA
Contact:

Re: The "Simple" No Math method ?

Post by Rich_Carlstedt »

Tony
There is a no math method as follows.
You only need to know the pitch of the thread, so it works for Metric or Standard threads with a 60 degree thread form.( An equallateral triangle)
Since the flank of the thread is the same length as the pitch , all you need is the pitch for calculations.
For example, a 40 TPI has a pitch of .025" ( One (inch) divided by 40 = .025).
So run the tool slide forward( with the toolbit in it) to get out backlash, and zero the (TS) dial, then run the crosslide forward till it touches the work, and zero it out (CS).
You now have your starting positions. All you need to do is advance the tool slide TS (.025 in this example) to match the pitch, and Bingo, its done.

A 16 TPI has a pitch of 1 Div by 16 = .0625
User avatar
Harold_V
Posts: 20248
Joined: Fri Dec 20, 2002 11:02 pm
Location: Onalaska, WA USA

Re: The "Simple" No Math method ?

Post by Harold_V »

Tony
There is a no math method as follows.
You only need to know the pitch of the thread, so it works for Metric or Standard threads with a 60 degree thread form.( An equallateral triangle)
Since the flank of the thread is the same length as the pitch , all you need is the pitch for calculations.
For example, a 40 TPI has a pitch of .025" ( One (inch) divided by 40 = .025).
So run the tool slide forward( with the toolbit in it) to get out backlash, and zero the (TS) dial, then run the crosslide forward till it touches the work, and zero it out (CS).
You now have your starting positions. All you need to do is advance the tool slide TS (.025 in this example) to match the pitch, and Bingo, its done.

A 16 TPI has a pitch of 1 Div by 16 = .0625

Yep, it's done, but where is it?

Is your compound rest (TS) dial calibrated on the diameter or the radius?

Did you start with a threading tool with a sharp point?

(or)Did your threading tool have the necessary flat for the thread pitch in question as it should?

Did you compensate the thread depth, allowing for the flat?

What was the diameter of the stock when you picked up with the tool, and if it was not proper, did you allow for the differential?

Is your compound set at 30°? If so, why? To avoid proper procedures? It should not be set at 30°.

If not, have you made the necessary adjustments for the error in feed because you are using a lesser angle?

Without measuring, you have no idea what you've accomplished. Maybe it doesn't matter, but if it does, how will you know when it's correct?

The theory of threads is best left to discussion. The cutting is best done by measuring. How else can you expect proper results?

Harold
Wise people talk because they have something to say. Fools talk because they have to say something.
Jacin
Posts: 1046
Joined: Sat Dec 21, 2002 12:14 am
Location: Near Cleveland, Ohio

Re: The "Simple" No Math method ?

Post by Jacin »

The theory of threads is best left to discussion. The cutting is best done by measuring......
Harold

Well that about sums it up perfectly. I would add that to make life easier the theory and formula be used to allow one to make "most" of the cut - purposefully stopping short and then measure the PD and next take a last finishing cut or two or three.

I think the combination of BOTH methods may be what would help us amateurs and even some of the "mildly experienced" out the most. Just like sneaking up on a cut. I true the part (light cut) then measure - reset my dials to "read" correctly. I make my cuts and ocasionally see if "dialing" say .030" or .060" or whatever will REALLY cut .030" or .060" or whatever (sometimes it does and sometimes it doesn't) When I am close I try to take a couple light cuts - each time seeing if the "dials" and real world results were at odds with one another or not - adjust as you go. Taking 2 or 3 light cuts and measuring each one is a VERY predictable means of achieving exacting results. Do I do this everytime?? No I don't but I do find myself doing it FAR more than I need to - force of habit - and I now very rarely "miss" my cut diension. I see threading (in general) being no different (unless you're talking about very tough materials where the DOC needs to be greater) - so in the general case the formulas - would be good to allow one to make several "known" cuts - and then follow them up with accepted measuring techniques - skipping the measuring - well you takes your chances then!! Me I like to know what I just made - scrap or otherwise [img]/ubb/images/graemlins/grin.gif"%20alt="[/img]

Happy chip making
John Garner

Back to Rich C

Post by John Garner »

Rich --

Your no-math approach is theoretically sound only if you are cutting the obsolete-for-over-a-century 60 degree Sharp V threadform with the compound slide "slewed" to feed the toolbit along the flank of the thread. (Which is the threadform specified in the original question.)

The analogous slewed-compound-infeed for the US Standard threadform is 3/4 x Pitch, and for both the Unified and ISO Metric threadforms 5/8 x Pitch . . . assuming a toolbit with a perfectly proportioned flat tip suitable to the threadform being cut.

As others have already pointed out, relying on infeed control to judge the completeness of a cut screwthread assumes that the individual error contributions from a small host of other sources sum to zero.

John
len
Posts: 421
Joined: Sun Jan 05, 2003 1:49 pm

Re: Back to Rich C

Post by len »

John,

Assumming one starts off with a known diameter workpiece, wouldn't using a standard thread gage to measure the thread be sufficient to approximate the final pitch diameter of a thread one is trying to cut? For precision work one should, of course, use best practices, but lacking the proper tools and/or knowledge to do it "right" doesn't mean to me that you shouldn't do it at all.

This discussion sorta reminds me of an item in John Muir's classic book "How to Keep Your Wolkswagon Alive; a Manual ... for the Complete Idiot", which was popular in the 70's. The item was how to take off the 36mm nut that held the flywheel to the crankshaft. The correct way, of course, was to buy a 36mm socket and do it right. Lacking such item, John suggested using a chisel to bang the nut loose. As he said in his book, you get to take it off and put it backt 3 times before you have to buy a new nut. Made sense to me back then, and still does today. The point of all this is that one often has to make do with what one has at hand. Theory quickly flies out the window in the face of necessity.

len
User avatar
Harold_V
Posts: 20248
Joined: Fri Dec 20, 2002 11:02 pm
Location: Onalaska, WA USA

Re: Back to Rich C

Post by Harold_V »

John,

Assumming one starts off with a known diameter workpiece, wouldn't using a standard thread gage to measure the thread be sufficient to approximate the final pitch diameter of a thread one is trying to cut? For precision work one should, of course, use best practices, but lacking the proper tools and/or knowledge to do it "right" doesn't mean to me that you shouldn't do it at all.

I'm not John, but I'll be damned if I'll let this one pass.

First, let me understand your question as posed. You suggest that one use a standard thread gage to measure the thread to closely approximate the thread diameter.

A standard thread gage is intended to be used that way, so, yes, that would be an acceptable method. My question is, if you consider your shop to be under tooled, where did you get all these gages? Why not use them and render this point moot?

If you're talking about using a "pitch" gage, I'm afraid you still need some swatting at the books. Pitch gages are just that, gages for you to determine a given pitch, and have nothing to do with gauging thread pitch diameters, although in some perverted way you could use one in that fashion. The next problem would be in overcoming the narrow pitch diameter tolerance as opposed to the variables in thread form. It's like using a yard stick to measure in thousandths. Sounds good, but it doesn't work.

Like it or not, the number of variables in threads make it impossible for any type of profile gage to act as a pitch diameter gage, and that's the way it is and will be. You can tap dance around this issue as long as you want, but when the music stops, measuring threads by some means, ring or snap gages, or thread wires or triangles, is the way you discover if your thread will fit, or not. That's the reality of threads, so now you can make a decision about "do I want to know it fits, or will I take a chance that it, perhaps, won't?"

One thing I will tell you, and it's guaranteed. If you thread by the formula, using geometry, the typical thread generated on a lathe will usually be well undersized. It fits, but is wrong. Now it's down to personal pride. Some of us have it, and in abundance, others have none, or very little. That's your call. Personally, I don't remove nuts with a chisel, and would be embarrassed to do so. I don't even use channel locks for turning nuts. Can't stand dog-eared fasteners and have been known to replace them simply because a wrench has slipped off and rounded the corners. If you don't mind rounded or otherwise chewed up nuts on an engine, or you don't mind using an undersized thread, that's your call. As long as you're comfortable making those decisions, what's the harm?

However, for those that want to know how to do it right, the information that has been provided by each of us that have worked in the real world is dead on, in keeping with typical machine shop practice, and mandatory in shops where QC is enforced. You come down on the side that best suits you and your needs. I'll not call the thread cops on you. (Though I may have then formed an unflattering opinion of your work!)

Harold
Wise people talk because they have something to say. Fools talk because they have to say something.
wlbrown

Re: Lathe threadiing math

Post by wlbrown »

TONY,
when you made your post, i bet you never thought you would
get this much feedback, and i guess you are more confused
than before. i can understand why. not many replies contained
a formula, only discussion of three wire system, pitch diameter,
and such. if you read my post, you will see i gave you a formula,
and here is another one that comes close to the first one.
divided .75/N, number of threads. in your case .75/16 =.046
this is the feed in on the compound.
good luck with your projects.
william l. brown
wright city, mo.
len
Posts: 421
Joined: Sun Jan 05, 2003 1:49 pm

Re: Back to Rich C

Post by len »

Harold,

I have gages for measuring thread pitch because they came with my tap and die set. Obviously, they can't be used to measure pitch diameter directly. But if a correctly formed thread is cut on a known diameter cylinder, I would hardly consider trying to determine the pitch diameter of the thread as "perverse", especially if the thread form is a sharp V.

Noone is disagreeing with you about the "proper" methods for generating and measuring accurate threads, but you seem to have trouble accepting the fact that some of us aren't striving for the perfection that you preach. Sometimes "good enough" is really good enough. And yes, sometimes you have to use a chisel to knock out a stubborn nut, not because you like to do it that way, but because you either don't have the $25 to spend on a 36mm socket, or because you would rather spend it on something else.

Sounds like you would have made a poor hippie. [img]/ubb/images/graemlins/wink.gif"%20alt="[/img]

len
User avatar
Victor_R
Posts: 688
Joined: Sun Jan 05, 2003 8:46 pm
Location: Taxland, Northern NJ

Re: "Back to" Harold #41359 & 41395

Post by Victor_R »

Harold, I love it. Way to go... SLAM-DUNK!
"The machines are gaining ground upon us; day by day we are becoming more subservient to them" ~ Samuel Butler (1863)
User avatar
Harold_V
Posts: 20248
Joined: Fri Dec 20, 2002 11:02 pm
Location: Onalaska, WA USA

Re: Back to Rich C

Post by Harold_V »

Harold,

I have gages for measuring thread pitch because they came with my tap and die set. Obviously, they can't be used to measure pitch diameter directly. But if a correctly formed thread is cut on a known diameter cylinder, I would hardly consider trying to determine the pitch diameter of the thread as "perverse", especially if the thread form is a sharp V.

It's hard to argue the point in principle, len, it really is. But in the real world, where one has to perform under the scrutiny of QC, you quickly discover that the theory doesn't work. I can visually check a thread that close, no gage needed.

Did you take note of my statement about the relatively narrow thread pitch tolerance? Small threads can be only a couple thou. The human eye can't discern anything less than about .003", assuming you have good eyesight.. How is your gage going to help you?

I've been in the shop for over 45 years and have seen it all. What my time there has taught me is that you take nothing for granted. We measure so we know where we are. There are times when that may not be necessary, and I'm the first to admit to that fact. It can be true, even of threads. However, it is poor practice.

There's a little method in my madness, part of which is in defense of my years of struggling to achieve a level of excellence on the machine of which I'm very proud. In fact, it was an experience with what I might call a friend that turned me into a raving maniac that you see today. This person, a self taught machinist, who had worked as an iron worker (structural) and had changed to jewelry manufacturing, was engaged in building a small fixture for his table top CNC mill, with which he was machining waxes for investment casting. He had no clue how to find center, so his indexing fixture had eccentricity problems. When I tried to explain to him how he could locate dead center, he refused to listen to advice that would have enriched him from that day forward. In his mind, it was not important, and to have listened to one that also claimed to be a machinist might just have been an admission to himself that he, maybe, wasn't one. Perhaps that's what some guys suffer with. Maybe they want to see themselves as the equal of the guy that has done it for years, secure in the knowledge that any damned fool can be a machinist.

Well, that's true. Any damned fool can be. But it takes someone with intelligence enough to know and understand that which is important and to know when and how to apply it to be a good one. Hacks are a dime a dozen, and even show up in the job shops, where they don't usually last long.

So then, my purpose here. I am doing my level best to have the reader understand that there are ways to do it right. Also to discern the difference, and why. I don't mind telling you that I am highly offended when someone comes along that doesn't have a clue and assumes that because they do something in a slip-shod manner, it becomes the way to do a job, and worse, they try to pass it off as the standard. A perfect example of the blind leading the lame.

It seems that those of us that have spent our lives perfecting a skill are treated as their equal. Anything we may have to say that doesn't comply with their preconceived notions becomes, somehow, irrelevent.

Lets put this shoe on the other foot, len. I have no clue what you've done in your life, how you made your living, whether you were (are) a well educated professional or a blue collar type, but lets assume you are a physician. How would you perceive a guy that insisted that scrubbing was unnecessary, that operating with a rusty blade was fine, so long as you could saw your way through the tissues, and anyone that can hold a knife can do surgery. If you stepped in with your expertise and tried to explain to this person that, while it could work that way, that it is not the accepted practice and there are other ways that offer more safety and likelihood of success, he looked for every conceivable way and reason to continue the course, secure in the knowledge that it didn't matter. That's what it looks like to me when someone does everything in their power to avoid doing it right, and expects a pat on the head for their cleverness. You want me to agree with you? Come up to the standard, don't insist on me lowering the bar. I can't afford it. More importently, I know better!

As I said, the method you choose has no bearing on me, nor my self image. It reflects strictly on you. You are the one that will use and live with your decisions. You, and only you, will make the decision to part with $15 to buy a set of p-d wires that will serve you reasonably well.
Noone is disagreeing with you about the "proper" methods for generating and measuring accurate threads, but you seem to have trouble accepting the fact that some of us aren't striving for the perfection that you preach. Sometimes "good enough" is really good enough. And yes, sometimes you have to use a chisel to kn! ock out a stubborn nut, not because you like to do it that way, but because you either don't have the $25 to spend on a 36mm socket, or because you would rather spend it on something else.

Guilty as charged! And I make no secret of it. I have little tolerance for poor craftsmanship, nor do I suffer fools gladly.

I've run machines long enough, and have worked around enough hacks, to know that it takes no longer to do things right than it does to do them wrong, and I'm sick to death of following up fools that refuse to do it right, but want someone to bail out their miserable asses when they fail. I do my level best to have people learn the principles and to do it right. Those that refuse I have little interest in supporting.

Early in my apprenticeship I was going to be fired for various reasons, including my immature attitude. I was only 18. I had one person show a personal interest in me, who convinced upper level management to extend my probationary period and to allow him to work with me. He took me aside and taught me, over a period of almost a year, to do it right, take it slow, check twice, cut once, etc., etc.. He made a machinist out of nothing. A literal silk purse out of a sow's ear.

This person taught me exactly that which I am trying to teach others, that there is only one way to learn things, and that is the proper way. With skill and knowledge, one can then make intelligent decisions about what will work and what won't. It's better to know how to walk before attempting to run, yes?
Sounds like you would have made a poor hippie. [img]/ubb/images/graemlins/smirk.gif"%20alt="[/img]

Yep! In spite of my over shoulder length hair, beard that has not been shaved off since 1964, my ear ring and my very outspoken independence, I'm not much of a hippie. I conform with the rules, drive the speed limit (I really do!) and I don't do drugs. [img]/ubb/images/graemlins/smile.gif"%20alt="[/img] I have been known to sip a bit of single malt, though. I like mine with water, heavy ice. [img]/ubb/images/graemlins/wink.gif"%20alt="[/img]

I make one hell of a good beatnik, though. [img]/ubb/images/graemlins/grin.gif"%20alt="[/img] (I'm saying this as I listen to some Errol Garner!)

Harold
Wise people talk because they have something to say. Fools talk because they have to say something.
John Garner

Back to William L Brown

Post by John Garner »

William --

You're missing a vital point: The flank length for a given Pitch of screwthread will vary with the threadform design. Your algorithm "Along-flank infeed = 0.75/N" (which is analytically identical to the "3/4 x Pitch" and "0.75 x Pitch" algorithms I cited in my earlier postings as Pitch is, by definition, equal to 1/N) is correct for the U S Standard threadform that has been obsolescent since 1949).

A different algorithm is necessary to calculate the flank length / along-flank infeed for 60 degree Sharp V that was specified in Tony's original question, and a third algorithm is necessary to calculate the flank length of Unified and ISO Metric threadforms.

Using the wrong algorithm to calculate the flank length is pretty much a lead-pipe-cinch way to cut the thread wrong.

John
Post Reply