Screwcutting on the lathe

Iain I've done a little drawing in cad today and your 0.92mm infeed at 30 degrees comes up slightly short of the minimum depth you'd need to reach the minor diameter for the nut so it would never go on, and it would require a 0.4mm flat to attain the proper flank width at that depth anyway. Obviously you don't want that because your thread is 1. not quite deep enough and 2. has no root clearance.

If your tip width was 0.25mm (10 thou) you could infeed 1.06mm and you would have a usable thread with good root clearance.

With a sharp point you need 1.31mm infeed, at 30 degrees to achieve the proper thread width at the pitch line. This is not very desirable because the sharp point makes for a much weaker thread.

Edited By Pete Rimmer on 04/12/2019 11:51:13

Right now I am totally at a loss. And apologies for the gap - a new mill got in the way.

I had another go.

The first thing i did was to compare the cutting tool against an M10 bolt (1.5mm pitch).

Apologies for the quality of the photo - it's one of many things I'm no good at!

by eye and even here, the tool matches the profile of the thread quite well.

Then I set up the lather to screw cut - centre in the headstock and all.

Then went straight in. None of this angled nonsense.

What I found was that at 0.9mm or so a nut would not go on the thread. this is the theoretical depth of the thread. I kept going in 2 though (0.05mm) increments. The first point at which the nut would go on was at a depth of 1.3mm.

Even at that it was a bit of a struggle. And when I ran a M10 die over it there was still swarf coming off.

So. It looks like I've misunderstood the various postings and I need to cut about 1.3mm deep not the 0.9mm I'd understood from reading).

Incidentally, applying O Level maths the depth of a 1.5mm wide 60degree angle triangle works out to be 1.299mm - roughly in accord with what I've found.

The other thing I found was that I was cutting in increments of about 2 thou (0.05mm) all the way up from about 0.8mm or the lathe would stall. Even at that I hard to run a little faster then I was comfortable with (250 rpm) to be sure to get the cutt - which meant more of a struggle to get it to stop at just the right place.

I remain confused about why ALL the books I've read on screwcutting seem to miss out the the depth calculation as it it were easy.

Any advice welcome - as always.

Iain

Iain,

Just a thought, did you reduce the bar diameter to about 9.8 mm to give yourself a bit of clearance on the bolt crests?

Tony

Metric threads are quite easy, at least if a calculator is handy and you have a good memory!

The depth of cut is always 0.87 × pitch, so M10x1.5 is

1.5×0.87 = 1.305mm (approx)

My usual mistake is forgetting my cross-slide is graduated in reduction of diameter, which is half depth of cut.

A confusion may be because properly formed metric threads have flattened tops an valleys which suggests a smaller doc than ×0.87 does. Maybe I'm lazy, but I just go in full depth and then finish off by flattening the tops slightly with a file as it spins. Also helps to give it a quick final polish with a coarse plastic pan scourer,

Dave

If you put a micrometer across the newly cut threads you will probably find the diameter has increased compared to the original stock diameter. There is usually some plastic deformation of material near the crests that pushes it out rather than perfectly cutting it. This is why you can buy full form threading inserts, they clean off this material. The problem is you need one to suit each thread size so are usually not worth the expense for hobbyists. It is also why the earlier advice to run a file over the new thread makes sense. The accurate way to measure threads is the three wire method which ignores problems with crests and roots but is probably over the top for you current needs.

The advantage of threading in the lathe is the thread pitch should be correct, the thread form should be axial to the stock and the thread can be made tight or slack as required. Cutting a thread with a die in a die stock with the material in a vice can result in wonky (technical term) threads. Using a die in a tailstock die holder may result in pitch errors if there is too much resistance to the die following the thread as it is formed. However once the thread has been formed in the lathe it can be polished with a suitable die in a die holder. If you have a split die start with it opened up and see if it removes some material. Try the nut on it and if it is too tight close the die slightly and try again. Repeat until you have a good thread.

Martin C

Well done. As you are finding, screwcutting is somewhat of an art and not just science.

There is always a burr, and as Martin points out some deformation, thrown up from the crest of the thread when screwcutting. This needs to be removed by running the corner of a flat file along the thread with the lathe under power. Keep the file moving at all times and run it along the thread several times. Otherwise, yes you will have trouble getting a nut to go on there.

Also, you should take a cut over the outside diameter of the blank before screwcutting so the thread OD is both concentric and slightly under the nominal size for clearance's sake.

As for your query about books and thread depth calculations, it's in all the old books I have. (Sparey, Duplex, GHT etc) Interestingly they usually quote a thread depth of about .64 x pitch for a 60 degree UNC thread. Which is way off from the modern recommendation of .87 x pitch for 60 degree metric. Bit of a mystery but I think the latter refers to a pointed tool bit with the job OD at the nominal maximum. In line with a pure right angle triangle calculation The old books I think assumed a radius on the tool tip and the blank OD starting out undersize for clearance. Hence less thread depth.

But once you have worked out what works for you and your lathe, stick with it and go for it!

Whichever way you go, the ultimate check is try and fit to the nut. Theory and measurements are one thing. Getting the darned things to screw together can often be something again.

Pretty sure Hopper is right.

An amateur thread cut with a 60° sharp V should be 0.89 x Pitch deep. But properly cut UNC/UNF/Metric threads aren't sharp V's. See footnote #

For practical reasons correctly formed UNC/UNF/Metric threads are flattened slightly top and bottom. Not allowing a sharp V at the base makes the thread stronger by removing a stress raiser. Slicing off the sharp top makes the thread easier to insert, less likely to jamb, and harder to damage accidentally. All good stuff, but tweaking the shape of the thread to meet the exact specification alters the dimensions slightly. The specification says the depth of cut from the top of the 'V' to the bottom flat is 3/4 of Pitch, which about 0.65 as per Sparey and friends. That means the cutter cannot be a sharp V, without special manuipulation. The depth of cut is measured from the outer diameter, which is later removed. Quite confusing!

Then choices can be made:

• For rough work, where the thread doesn't have to be a strong as possible, it's reasonable to cut a sharp 0.87 deep V thread. It will fit. But this isn't a 'proper' thread.
• If the thread specification matters, better to cut 0.65 deep with a V cutter. The resulting thread won't fit, but it can be finished to size and shape with a die. The partial thread cut by the lathe is done to minimise wear and tear on the die rather than to produce a fully finished thread.
• It is possible to produce flat bottomed threads with a V cutter, see the angled tool-post technique rather than plunging straight in.
• If a properly shaped carbide or HSS thread cutter is available, cut to 0.65.

Also bear in mind Martin's point about plastic deformation. Machining often deviates somewhat from theory because tools wear, materials squidge, everything bends, and the operator makes small mistakes. Most of the time in a hobby workshop, I feel it's not necessary to meet industrial standards, for example, a sharp V is 'good enough'. Exactly how a job is done is often a judgement call, all part of the learning process. We'd get bored if it was easy!

Dave

# The thread form of all three is identical - no difference in shape, they differ only in pitch, so the same sums apply in this case to both Imperial and Metric.

Edited By SillyOldDuffer on 10/02/2020 10:18:39

Don't they just. What a great description

Edited By David Standing 1 on 10/02/2020 10:34:28

A friend of mine who restores Classic bikes, mainly BSA. Has had a lathe for over 35 years. A Colchester I think. Huge monster. His attitude is why would I ever want to screwcut when I have all the taps & dies.

But I can see the satisfaction of having a go at it. Something on my list for the future. Have a chimp to help me now so what could go wrong.

Steve.

Your mission is to educate the poor deluded bodger! He's wasted a Colchester for 35 years, be gentle with him.

Even owning a complete set of taps & dies there's advantage in screw-cutting with a lathe. I already mentioned the value of reducing wear and tear on dies by using a lathe to cut a partial thread as a guide. Another important reason is that lathe cut threads are accurately aligned; they don't wander as taps & dies are prone to do.

There's more - lathes can make any thread pitch from the gears it has available, and thread non-standard diameters, and make taps and dies. After being used to true up the face, etc, etc.

On second thoughts, don't be gentle. The chap has an attitude. Off with his half-nuts, he'll never miss 'em!

Dave

Ha ha good one.

Steve.

I've enjoyed reading this thread having been taught single point threading by an apprentice instructor in the 1980s and then come back to lathe work in the past couple of years screwcutting is definitely more tricky than you plan for... for all of the reasons stated.

Regarding getting the feed to stop in the correct place without crashing, yes running in reverse might be an option but only if your workholding device doesn't attach to the spindle with a screw thread. Mine does.

One option I have used myself and will chip in here is to use a handle to turn the spindle manually for the last few threads. I've used this myself and it makes it dead easy to cut a thread up to an exact stopping point - with the handle you can feel when you reach the end with no problem at all.

My 'handle' is nowhere near as sophisticated as those I've seen elsewhere. I bought some hex steel bar from an oddments bin, bored and threaded this with the 32 TPI thread to fit the back end of my spindle. The 'handle' is just a bog standard hex socket and a ratchet handle from my socket set. Works a treat.

I must be the odd one out, I don't find screw cutting difficult at all. I picked it up quite quickly and I actually really enjoy the process. I'm no expert though and its probably because I don't bother getting bogged down with all the technical details. I've always cut threads with the cutter at 90 degrees to the work and just go slow and steady, I use the handwheel graduations as nothing more than a guide. Trying to make accurate measurements with them is ridiculous in my opinion, especially when you have backlash, deflection and machine flex to consider. Just keep checking the thread with a thread guage or by screwing a nut on until it feel right. Very few of us need perfectly formed nasa grade threads! Keep it simple.

I want to chip in here. Get it chip , no ok. The strong point I want to make is something dad said to me when I was younger. A lot younger. Was because I was doing so bad at school he asked me about the history lesson's & said do you enjoy them, I said no. Do you like maths, no, do you like anything I said yes. What is it. Metalwork, I love it. Then that is something you will be good at, because if you don't like it you will never be any good at it. If you love it you will excel. MMMM if I could just find the on off sw for the lathe. I have some 33 34 change gears for the myford & have just got to have a go at metric on an imperial. I am sure it will all end in tears but after reading this thread I must at some point have a go.

Steve.

Hollowpoint,

You are a true hobbyist - not bothered about time and simply enjoy getting to the end product.

(a thread in this case).

Percy Verence was good at what he did, for that very reason. He always managed to get the job done eventually.

Lots of things are akin to building a house with a pack of cards. If you are missing any of the layers, the top ones will collapse - or not be supported properly, thereby making it difficult to finish the job.

If you miss out on the basics, the harder parts can become overwhelmingly difficult. That is the way I look at maths, for instance. If Percy had cried off at the first hurdle he would not have achieved as much as he did.

Today I was making a part for the later version of the Worden tool & cutter grinder (mine is over 15 years old). The part called for an accurate 1/4" x 40 TPI thread 1" long, so i screwcut it to a smidgen oversize and finished it off with a die. This gave me a thread which was truly concentric with the part, had an accurate thread--form, and was a slightly tight shake-free fit in a tap-threaded hole - which was exactly what I was trying to achieve.

Edited By Bob Unitt 1 on 12/02/2020 19:22:23

I'll take that as a compliment 🙂

I think we (engineer types) are all guilty of over complicating things at times. I know I do. This is when we get frustrated. Having a go and making mistakes is the best way to learn anything!

Having a go and making mistakes is the best way to learn anything!

Would agree for most things, but there are a few things I might want to get right first time! Bomb making and parachuting might be on that list - not that In would want to do either of those at my age!

Thanks for all the comments guys.

It looks like the weekend before I get back in the shed, but I will have another go at a 10mm bar and then work on an internal thread to match (shudder). If that all goes smoothly, I will thread the piece that I actually want to make!

Iain

Iain you're making it hard for yourself turning a thread on such a small diameter for a practice piece. A larger diameter thread would be a lot easier.