Metric Thread On Imperial Lathe

[doubleboost]

Hi
I have been buisy making a ER32 collet holder
This link should take you to a screw cutting video
Sorry about the geordie accent

[DavidA]

Nice demo.  I've had to do exactly that on numerous occasions at work;  but without the variable speed reduction to help. Always was a bit nerve wracking as the last turn approached.

Dave.

[Jonny]

Nice video.

Precisely why i sworn blind i would never buy another lathe that wasnt dual metric and imperial, no faffing around with change wheels.

Had to do similar at work with just the on/off switch and couldnt alter the speed, so used to revolve by hand and power it back.
Luckilly the Harrison i can disconnect lead screw wind carriage back and re engage lead screw again any where for most metric. I try to avoid shutting motor down as it blows a 13a fuse 1 in 3 startups else would wait for motor to stop, start in reverse etc.
Also would scare the crap out of me without a clutch, spindle stop etc.

[Rob.Wilson]

  very good clear video John 

I nass  what yu saw tarkin aboot 


Rob 

[Bluechip]

very good clear video John  

I nass  what yu saw tarkin aboot  


Rob  

I pasted  that into Babelfish, and translated into German ... look what I got

Nass I, welches yu tarkin aboot sah    

No better ... but certainly no worse, will have to try Mandarin or summat next ..

我nass什麼yu看了tarkin aboot   

Ah, that's fixed it. Why didn't you say that in the first place ????     

Dave BC

[Bogstandard]

Very good instructional vid John.

All you need to do now is make a swing up toolholder, and halve everything that you had to do.

But honestly, can't fault what you showed.


Nice one


John

[doubleboost]

Thanks John
I will have to make one
I should have videoed the other end
1 i/2 inch 8 tpi internal that was great fun 

[Ned Ludd]

Hi There,
I like your video clear and concise as all training videos should be.
I hope you will forgive a small criticism, have you not set the top/compound slide to 29.5 on the scale? Surely it should be 29.5 to the lathe axis, which should read 60.5 on the scale? If I have misread the situation I apologize.
Ned
EDIT It should be 29.5 to cross slide not lathe axis but otherwise info is correct

[Blade]

Nice video!

[Bogstandard]

have you not set the top/compound slide to 29.5 on the scale? Surely it should be 29.5 to the lathe axis, which should read 60.5 on the scale?

You are quite right Ned, but there are actually two ways of doing it, the way John has shown, which gives much better control on the depth of cut, and to me, if doing super fine threads, I would use, but only with 60 deg thread forms, the sidewards feed is a little too much for 55 deg or lower threads, and you risk actually cutting the threads away and so going undersize on the OD.
Or the normally accepted way, which you mentioned, and gives a more aggressive depth of cut, but less sideways feed, and is the way I would use most of the time.

So if you are not into the fine complexities of single point threading, you are quite correct

In fact, I have marked up my cross slide with datums on all main points of the compass, so that whatever type of threads I want to cut, both left and right hand, I can set up the degree scale for the types of threads and tapers I need to cut, as the scale on my compound only goes to 45 degs either way of zero, and so I can't set it to the required 60.5 for 60 deg threads against the normal datum mark.

John

[andyf]

have you not set the top/compound slide to 29.5 on the scale? Surely it should be 29.5 to the lathe axis, which should read 60.5 on the scale?

You are quite right Ned, but there are actually two ways of doing it, the way John has shown, which gives much better control on the depth of cut, and to me, if doing super fine threads, I would use, but only with 60 deg thread forms, the sidewards feed is a little too much for 55 deg or lower threads, and you risk actually cutting the threads away and so going undersize on the OD.
Or the normally accepted way, which you mentioned, and gives a more aggressive depth of cut, but less sideways feed, and is the way I would use most of the time....

John

John, I'm feeling a bit stupid, because I don't understand that   . I can see that it gives finer control over the depth of cut. But I should have thought that increasing the depth of cut using a topslide set at just over 30 deg to the lathe axis (rather than just over 60 deg) would result in a thread with its LH flank correct but the RH flank wrong - a lopsided 90 degree thread. If it's not too much trouble, would you explain a bit further so I can grasp how it works?

On the subject of threading, it might be worth mentioning G H Thomas's method of getting to the right depth when using an angled topslide, for anyone who doesn't already know it:
1. Feed cross-slide forward until tool just touches the work (nip cigarette paper if desired). Set both the cross- and topslide dials to zero.
2. Move saddle to the right, so tool is clear of the work, advance the cross-slide by the depth of thread, and reset its dial to zero.
3. Withdraw the tool using the topslide, move saddle to the left, advance topslide until tool just touches the work, and then start cutting the thread in the usual way, applying depth of cut using the topslide. Unless you have a swing-up toolholder, pull the tool clear of the work with the cross-slide when traversing back, and re-advance the cross-slide so its dial reads zero before applying more depth of cut with the topslide for the next cut.
4. The thread depth is correct when the topslide dial has got back to zero again. Of course, if the thread is a deep one, the topslide dial may need to go round more than once before you get there, but when passing through intermediate zeros it will be obvious that the thread is only part formed.

This avoids resorting to things like trig or Pythagoras; I always was a bit of a duffer at sums.

Andy

[Bogstandard]

Andy,

I am not putting it off, but I am having trouble with my comp and it might be a few days, but I will definitely get back to you on this one.

I have just lost around ten years, about 300GB, of engineering archive material, and I am trying to get it back. Not my fault, my archive drive (almost new) has decided to say that it is an unformatted disk. So I may be offline for a couple of days.


John

[Pete.]

On the subject of threading, it might be worth mentioning G H Thomas's method of getting to the right depth when using an angled topslide, for anyone who doesn't already know it:
1. Feed cross-slide forward until tool just touches the work (nip cigarette paper if desired). Set both the cross- and topslide dials to zero.
2. Move saddle to the right, so tool is clear of the work, advance the cross-slide by the depth of thread, and reset its dial to zero.
3. Withdraw the tool using the topslide, move saddle to the left, advance until tool just touches the work, and then start cutting the thread in the usual way, applying depth of cut using the topslide. Unless you have a swing-up toolholder) pull the tool clear of the work with the cross-slide when traversing back, and re-advance the cross-slide so its dial reads zero before applying more depth of cut with the topslide for the next cut.
4. The thread depth is correct when the topslide dial has got back to zero again. Of course, if the thread is a deep one, the topslide dial may need to go round more than once before you get there, but when passing through intermediate zeros it will be obvious that the thread is only part formed.

This avoids resorting to things like trig or Pythagoras; I always was a bit of a duffer at sums.

Andy

What a great method - I for one didn't know that!

[RichardShute]

have you not set the top/compound slide to 29.5 on the scale? Surely it should be 29.5 to the lathe axis, which should read 60.5 on the scale?

You are quite right Ned, but there are actually two ways of doing it, the way John has shown,
<....>

John

John, I'm feeling a bit stupid, because I don't understand that
<....>

[/quote]

Nor can I understand it, given the good description at 3:40 into the video, which relates one of the normal methods very nicely. With the compound slide set as shown the resulting thread form will be like the pic below. The purple is what you would be hoping for, but the saw-toothing on a right flank and reduced peak on a left is what you will get.
It's not a question of a fine thread or otherwise, it's a geometry issue, just at a different scales.

You can leave the compound parallel to the spindle axis and apply half the infeed to that and the resulting tool path is near enough the same as with the rotated compound method (arcTan0.5=26.6deg from the cross slide), but that is not the description here.

Richard - confused

[doubleboost]

Hi
Richard
I hope these scans will help
I dont fully understand how it works but it is the method i have always used.
I should have stated on the vid that the last cut was made on the cross slide (only about 5 thou )
I have looked at the thread through a magnifying glass and it is as good as the thread on the nut.





Hope this helps
John

[RichardShute]

Thanks John,
that is the method you described very well in the video and I agree it is one of the widely accepted methods. The query/confusion comes because in the video, you had the compound set so as to have the angle shown in the diagrams as 27.5deg actually set to 61deg.

You seemed to have set the compound at 29 deg from the spindle axis as I tried to show in my sketch, not 29 deg from the cross slide axis as shown in your scans. Maybe I've got the wrong end of the stick because your threads look fine.

Richard

[doubleboost]

This scan is taken from the boxford hand book
the angle is 29 deg

[RichardShute]

Hello John,
Maybe I was making things worse, I'm not worried about whether it's 29 or 29.5 or 27.5 deg, but where the degrees are measured from.

From what you are saying, when your compound slide is parallel to the cross slide, the degree scale must read Zero. On my lathe the compound reads Zero when it is parallel to the spindle axis. There's the difference/confusion.

Richard

[doubleboost]

I will go have a look

[Pete.]

I agree with Richard. It all depends on how the scale is marked up on the lathe. In the video, the compound is angled 30° off parallel to the spindle axis, where it should be 30° off perpendicular. My lathe is the same I have to set the scale to 60° (or actually 60.5) for a metric thread. The way it's set in the vid will give a 91° angle thread rather than a 60.

[doubleboost]

That means the boxford hand book is wrong

Regards
John

[BillTodd]

I agree with Richard. It all depends on how the scale is marked up on the lathe. In the video, the compound is angled 30° off parallel to the spindle axis, where it should be 30° off perpendicular. My lathe is the same I have to set the scale to 60° (or actually 60.5) for a metric thread. The way it's set in the vid will give a 91° angle thread rather than a 60.

Yes, On my Hardinge it is 59°

[doubleboost]

I am really confused now
Interesting stuff though
I would go and cut some threads if it were not below freezing in the shop

[BillTodd]

I think the difference (and thus confusion) is down to how the trailing side of the thread form is to be cut.

If the top-slide angle is less-than the the thread angle ( i.e 29.5° for a 30° thread) then the trailing edge is cut by the back edge of the tool.

If the angle is greater than the thread angle then the trailing edge is cut by the tip of the tool, (as the back edge clears the face) and is thus defined by the angle of the slide.

If the slide is set to exactly the thread angle, then the cutter just scrapes the edge.


Bill

[RichardShute]

That means the boxford hand book is wrong

Regards
John

Well it's not really 'wrong', just rather ambiguous. It says '...set at an angle of 29 deg...' which a legalese pedant would argue is not quite the same as saying 'set the compound slide so that the scale reads 29deg.'
It is very poorly worded, but at least we've got to the bottom of it.

Richard

[RichardShute]

<....>


If the angle is greater than the thread angle then the trailing edge is cut by the tip of the tool, (as the back edge clears the face) and is thus defined by the angle of the slide.
<....>
Bill

Bill,
your first drawing is essentially the same a mine. I showed the saw toothing as a result of each successive cut taken - rather coarse I accept, but it was intended to accentuate the effect. I agree with your other two drawings, I was just too lazy to bother....

Richard

[BillTodd]

your first drawing is essentially the same a mine.

Yes it is, Richard. Sorry I missed that first time through

Of course, all this means is that me and every other Hardinge owner (the ones that follow the manual anyway) have been cutting slightly incorrect threads  (albeit with the greatest of ease).

Bill

[Ned Ludd]

It is regrettable that so few books show that what they mean by "1/2 the thread angle" does not in fact relate to the markings on the cross slide's quadrant. The half angle must be subtracted from 90 and that figure set on the quadrant. This is one of the problems of self taught machinists working in their sheds alone. They have no one to point out elementary errors, whereas those lucky enough to have mentors are blessed.

I must also point out it is a bit of a miss-perception that going straight in (not the set over top slide method) loads the tool on both sides. It does no such thing, once the cut is under way. Cutting a thread is no different to an ordinary sliding cut, just deeper, think about it! For an ordinary RH thread, the tool will only cut on the left side. The only time a threading tool would cut on both sides is when you do a plunge cut. As the old saying goes "unless you know different".
Ned

[Pete.]

Actually now I think of it what you produce using a 30°-off-parallel angle for the compound is a saw-tooth pattern on the right flank of the thread because the angle you're advancing at is shallower than the angle of the cutter. My CAD skillz are lacking so I'll knock-up a quick pencil-sketch.

[Pete.]

Ned I don't agree, if the tool is not loaded both sides what is cutting the right flank of the thread? I see feeding directly in exactly the same as plunge cutting except that you are continuing the plunge cut along the length of the thread, a 'running plunge cut' for want of a better term.

Anyway, here is my crummy sketch depicting the result of feeding in at 60° off perpendicular.

[BillTodd]

Here's a 1" 12tpi thread I cut the other day (@ ~400rpm with a full depth insert in BMS )

You can see the striations caused by the each pass. However, although it is cut at 31° (59° on my top slide), there are no obvious steps on the trailing edge (right-hand side).

[edit] Actually, now that I can see the tips clearly, it looks like I've over done the in-feed (or picked up the wrong insert??) . As it happens I'm going to have to slice off and re-cut the thing as a dual start thread (i.e. 6tpi) as soon as I get the appropriate change gears for the lathe.

Bill

[RichardShute]

Bill,
as you say, it looks fine. A 1 degree error is a near miss, but we were initially talking about a 31deg error and that would make a significant difference.

The saw toothing may be less obvious depending on the tip radius of the tool, but the reduced height of the thread profile would still be the same as each successive cut removes a little more of the crest with the trailing edge of the tool.

Richard

[Bogstandard]

I think you are all missing the major point here.

It really matters not one iota at what sort of angle you are coming in at. Half the thread form angle minus 1/2 degree from the cross feed motion is observed to be the optimum, as at that figure, you are keeping the trailing edge fairly clean cut.

But you should not be worried about what is happening on the trailing edge, purely because you should always finish your threading cut with a minute feed of the cross slide, which then cleans up perfectly both faces of the cut thread, as long as you have set the cutting tip up correctly with the gauge in the first place.

I stated that I use the compound at a shallower angle for cutting very fine threads, which is what I do.

When you set your compound over, you can control the depth of cut much better than by using the cross feed, and by setting it at say 5 degs from parallel to the spindle, you can use the feed of the compound to take parts of a tenth off the diameter of a bar during normal lathe turning operations.

I use that method, but using a slightly larger angle, so that I have much better control of the depth of thread when cutting fine ones.

Maybe not in the books, but it was shown to me many years ago by an old timer as a method to get superior results, and I adopted it. The final tiny cross feed takes care of any 'sharks tooth' effect on the trailing edge cut.

Everyone has their own methods of cutting single point threads, but there is no need to argue over it or shove it down everyone elses' throats because you think their method is incorrect.
Make your methods known by all means, and if someone wants to take it on, then they will do, but for other people who have their own methods of doing it, and if it serves them well, and are quite happy to do it that way, then certainly let them carry on using it.

Nothing is written in stone, in fact for thread cutting, I think there are more conflicting words of text than for any other machining operation, so it has to be a personal decision as to which commandment you follow.


Bogs

[Ned Ludd]

Hi Pete,
As I said "once the cut is under way" how can the tool cut anything on its right hand edge when it is moving sideways to the left. If I am wrong, wouldn't be the first time, the only cut on the right of the tool would be the small incremental in-feed, while the left hand edge cuts full depth, I would not describe this as "crowding" would you?

As Bogs very wisely says, nothing is set in stone about thread cutting. What matters is that you can produce a decent thread, first and foremost. With more experience you can vary your methods to suit your different needs or tooling. There are Pros and Cons for all the various different methods of single point thread cutting. I have never seen a fully comprehensive treatise on single point threading, but I am sure it would run to hundreds of pages and even then some chap working quietly by himself in a shed, would have yet another different way of doing things. One of the best books for the home machinist is the third in the Workshop Practice Series, "screw cutting in the lathe" but it is now sadly a little outdated, even so it still makes a good primer for those new to the subject. If it were to be up-dated it would have to include some modern developments like Bogs' own type of self lifting tooling.

What is important is that instructional books, or other teaching methods, are not open to misinterpretation by an averagely mechanically minded person.
Ned

[andyf]

I think that I was misled by the video. The camera angle was such that I mistakenly thought the topslide was set at a much greater angle than 29.5 deg to the cross-slide axis.

Andy

[RichardShute]

I think that I was misled by the video. The camera angle was such that I mistakenly thought the topslide was set at a much greater angle than 29.5 deg to the cross-slide axis.

Andy

That is the nub of this whole discussion.

In the video John gave a clear description of the well documented rotated-compound-slide method and says the cut is taken on the left of the tool with the right of the tool just scuffing the rear face to keep it clean. BUT... when the top slide was rotated in the video, it was indeed rotated by 29.5 deg from the spindle axis as John agreed in a later part of the discussion.
The inevitable consequence of this is that the top of the thread crest is gradually removed with each successive cut as the thread is formed. In the end, although the root diameter is correct, the crest of the thread is undersize by roughly half the thread height as shown in my diagram and Pete's and implied by the last of Bill's.

It does, surely, matter a lot what the approach angle is, if the result is a significantly truncated thread. Using a low angle to get a fine feed is a well established technique, but in plain turning there is no concern about preserving a particular tool path, in thread cutting, by definition, there is - it's the thread.

Ned, I was wrong to say the crest is removed by the heel of the tool, that is not the case, the top of the thread is taken away by the tip on successive passes. The tool could only cut on the right as well as the left if the compound angle were less than half a thread angle from the cross slide, but not at the 30deg from spindle axis we are discussing. Sorry for mis-leading you.

doubleboost, I would be interested to know what is the OD of your finished thread. If you started with 40.00 and cut a 1.5 pitch thread, with the compound set as shown in the video I would expect you to have lost about 0.6mm per side so the finished OD would be about 38.8, depending on the tool tip radius.

Richard

[Pete.]

Hi Pete,
As I said "once the cut is under way" how can the tool cut anything on its right hand edge when it is moving sideways to the left. If I am wrong, wouldn't be the first time, the only cut on the right of the tool would be the small incremental in-feed, while the left hand edge cuts full depth, I would not describe this as "crowding" would you?

Yes I would.

Let me start by saying that I am a novice machinist and as such have been and am happy to be corrected if something I say is not right.

When you directly in-feed, you take the first pass and the tool cuts both sides. The leadscrew moves the tool along it's path continually cutting both sides up to the end of the thread. When you set up for the next pass, the leadscrew moves the tool along exactly the same path as controlled by the leadscrew, but the tool is fed in extra depth, so the tool is still cutting both sides but deeper.
Look at it another way, imagine you were cutting a dead-square thread with a full-width square tool. After the first pass, you return and feed the tool directly in using the cross-slide and take another pass. Will the tool now cut the thread deeper by cutting on the end of the tool, or wider by cutting on the side of the direction it's travelling?

As Bogs very wisely says, nothing is set in stone about thread cutting. What matters is that you can produce a decent thread, first and foremost. With more experience you can vary your methods to suit your different needs or tooling. There are Pros and Cons for all the various different methods of single point thread cutting. I have never seen a fully comprehensive treatise on single point threading, but I am sure it would run to hundreds of pages and even then some chap working quietly by himself in a shed, would have yet another different way of doing things. One of the best books for the home machinist is the third in the Workshop Practice Series, "screw cutting in the lathe" but it is now sadly a little outdated, even so it still makes a good primer for those new to the subject. If it were to be up-dated it would have to include some modern developments like Bogs' own type of self lifting tooling.

What is important is that instructional books, or other teaching methods, are not open to misinterpretation by an averagely mechanically minded person.
Ned

Wise words indeed but there should always be room for someone to disagree with anything that's said.

With the very greatest of respect, I actually don't agree with Bogs that it's acceptable for the compound angle to be greater than half the included thread angle especially when it's double because I think it will produce the saw-tooth form as shown in my sketch. I don't see how a few thou cleanup cut is going to be enough to turn a 90-degree included angle thread into a 60-degree one especially on the OP's 12TPI thread, but as I said I'm a novice machinist with no more than a handful of single-point threads to my name so there might be something I'm missing there. I certainly won't stamp my feet and get all silly if I'm proved wrong, but I'd like to BE proved wrong rather than be told not to argue when I think I'm not.

EDIT: It was Bill Todd who cut a 12TPI thread, the OP cut a 1.5mm pitch thread for his collet chuck.

[Ned Ludd]

Hi Pete,
Argue away, without discussion how is anyone going to get out of the rut of narrow thought?

I, too, have to be convinced about cutting threads at shallow angles, as I wrote in an earlier post I would have thought that the last third of the cut would have to be in-feed, to correct the thread form. OOPS, I just checked back to see exactly what I posted, no post. I wrote a post, I know I did,  but it seems it did not get through and I did not check, silly me. 
Ned

[Ned Ludd]

Hi Guys,
My apologies, I must have saved my post and not sent it. The following was written as a reply to Bogs and should have appeared on page one.


Hi Bogs,
I do have a little experience in thread cutting, and the only way I can see the method in the video working would be if the last third of the cutting was of the "straight-in" method. By doing it that way, after the bulk has been taken away, the correct thread form could be produced by the insert. Though having said that, as he is using "proper" inserts, why not just go straight in and use the insert as it is designed to be used?

While on the subject of threading, I have been experimenting with the swing up tool. I think yours is a stiffer and superior  design than the German chap's. To be fair to him, he was only cutting very fine threads not 1/2" Whit!  I have tried JohnS's with die head chasers, but on some metals the loading is, I find, too much. I am currently trying one using triangular carbide threading inserts, which shows promise. The next experiment will be with the Oz design of tangential threading tool in your style swing up holder.
So far I have found the best method of getting a clean cut thread is to use the Oz cutting tool shape in my own design holder, using the offset top slide method but with a cross slide stop for quick action.
Ned

[doubleboost]

Hi
I made another chuck (for the rotary table)
I used the same insert but went straight in this time.
The diamiter before machining the threads on both pices was39.8mm.
The finished thread measures 39.7 on both jobs and the threads look identical.
I proberly lost the 0.1 as i polished the finished thread with emery tape.
Regards
John

[BillTodd]

Well, I braved the freezing temperature in my garage to cut a thread (20tpi with a sharp V tool) into a piece of scrap HRS at 30° (to parallel) and again at 61° (I'd normally use 59°)

The result was a failure as 30°, partly because of the Hardinge's fast retract top slide tore up the thread as it was pulled out, so requiring me to use the cross-slide to retract (thus ballzing it up). As expected the tool was cutting on both flanks, which doesn't make for a nice looking chip. You may be able cut a thread at this angle, but frankly it doesn't seem or feel 'right' and you'd get a better result feeding straight in with the cross-slide

At 61° the tool was again cutting on both edges which creates a squiggling tearing chip that does not compare to the nice curl at 59° that I'm used to. I will try it again (when it's warmer) with a full-depth insert to see if it make any difference.

Bill

[doubleboost]

Hi
Bill
One of the benifits of using the compound slide is that the tool is wound back on the cross slide (boxford do a threading stop for this very reason).
Like you say it is a bit cold to play in the shop  bad enough at work
John

[BillTodd]

Like you say it is a bit cold to play in the shop 

I'm only half Geordie, so I'm not as well adapted to the cold; -2° is all I can cope with 

[doubleboost]

off topic but it is christmas

[BillTodd]

geet . Aa'd not seen tha before

(Translated by http://www.geordie.org.uk/)

Bill

[RichardShute]

Hi
I made another chuck (for the rotary table)
I used the same insert but went straight in this time.
The diamiter before machining the threads on both pices was 39.8mm.
The finished thread measures 39.7 on both jobs and the threads look identical.
I proberly lost the 0.1 as i polished the finished thread with emery tape.
Regards
John

Thanks for taking the trouble John, I'm very curious to know how this comes about. One aspect I had not included in the sketch I did before was the shape of a real thread which has a clipped top and a radiused root, both of which reduce the effect, but I still can not see how your result comes about.
Here is a notional 60 deg thread with no flat top and no radius:


This is a single groove made with a 60 deg tool fed in at 30 deg from the spindle:


and this is two adjacent grooves on a 1.5mm pitch, showing the clipping at the top of the form:


That makes no allowance for the flat top or root radius and is the form I was using previously. Not strictly accurate I know, but it domonstrates the point of discussion.

This is a true thread form with a flat to a depth of H/8 and a root radius at the full quoted depth of 0.92 (both figures taken from a Zeus book):


This shows the top again clipped, but by a reduced amount of 0.33 because part of the top is already removed by the flat and the root radius also reduces the amount removed.

Can anyone shed light on why this appears not to be reflected in the thread actually cut.

Richard

[andyf]

I’ve been puzzling over the remark which Bogs made before his data got lost, and have come up with these approaches to “going in at an angle”.

Method 1. The diagram below shows how I normally set up for threading, with the topslide set at 60º to the work and at 30º to the cross-slide (actually 60.5º or 29.5º, but I’ve rounded them off). Depth of cut is applied using the topslide.



Method 2. The diagram below shows the set up for threading when applying depth of cut with the cross-slide, but also shifting the topslide (and thus the tool) towards the headstock a little between each pass to produce the angular feed effect. The rule of thumb seems to be that the topslide is shifted left half the amount by which the toplide has been fed in.


Method 3. The final diagram below shows the situation which seems to have caused controversy.  The topslide is at 30º to the work and at 60º to the cross-slide. This set-up could be used to simplify the “cross-slide parallel to the work” method. Trigonometry shows that advancing the topslide when it is set at this angle will advance the tool into the work by half the amount indicated by the topslide dial. So in this case the cross-slide and topslide can be advanced by the same amount as shown by their dials to produce the same result as Method 2 with its parallel topslide. This avoids the bother of remembering by how much the cross-slide has just been advanced and then advancing the topslide dial by half that amount. Each dial is advanced by the same amount in Method 3.



Am I on the right track. or have I got myself completely lost?


Andy

[Ned Ludd]

Hi Andy,
What an interesting idea, and  yet another method. 

Going back to the original post, the following You-Tube was pointed out on the ME site today. Alright, it is yet another video on threading and perhaps it is a touch long, there are five videos for the one thread, but I always like to watch others working. Oh, and it does seem a bit like an advert for tips and oil.

Ned

[Bogstandard]

I think you are going into too much theory, and I would suggest you read my very first statement on it.

if doing super fine threads

If you are going a lot deeper, then you will run into all sorts of problems, as shown above.

There are reasons for cutting on the leading face, and that boils down to removing metal without leaving behind the raggedy a**sed threads you usually get when using the straight in technique. What I actually do for very fine threads is to rough it out, without going too deep (hence the shallower angle, which gives greater control of the depth) then use a final straight in approach to clean the back edge up, with maybe one or two tenths infeed. I hope all that reads correctly, I've only just got up.

I find it works for me, but in theory, it shouldn't work. But as with a lot of things, practical doesn't always match theoretical.
That is why I hate it when people start throwing formulae around, it is just too confusing for the laymen who read such posts as ours, it makes them feel so inferior, that they never even try to do something out of the ordinary.

As I stated before, that was what was shown to me, where that old chappie got it from, I have no idea, but somewhere in the past, a machinist must have tried it out and found that it worked for him as well.

But I will say one thing, my tiny threads are always as smooth as silk.

I think where a lot of people go wrong is that they are only looking at standard sizes when threading, say 1/4" diam x 40 TPI. But in fact, you can just as easily have 3ft diam x 40TPI, exactly the same thread profiles, but on a much larger diameter. The critical part is getting that very fine thread to work with it's female counterpart.

Anyone who has visited my shop will see that I can get up to all sorts of weird things, but unfortunately, at this time I cannot show you anything, maybe next year after the patents are completed, but that also includes some strange  but very fine threads, all single point cut.

I hope Peter (HS93) doesn't mind, this is a job I am attempting to finish off for him, but here is the sort of thing I am on about. Making a few new blades for a model steam boat variable pitch propellor.

The hub is less than 1" diameter and is full of bevel gears, and the threads have to be a spot on match for the ones already cut in there. Because if they aren't, because of the very tight tolerances on the unit, the whole lot would sieze up. I have made five stubs, just in case I bugger up on the machining to follow.





So as I was saying in a previous post, if you find a method that works for you, use it. If you like what you see, try it, but don't always expect it to be right for you.


Bogs




 

[andyf]

Bogs: I think you are going into too much theory.....

You're right, of course. But it's so cold in the shop that sitting indoors and theorising is all I feel like doing just now. Of course, what really should be doing is sitting indoors finishing off my tax return, but that's an even less attractive proposition than a sub-zero garage!

My thanks to Doubleboost for his video. I wish I had a 127T gear to get exact metric threads out of my imperial leadscrew, but there isn't room for one in Mod 1, so I would need to use a smaller Mod for the 127T and whatever it meshed with. After a lot of calculations involving my standard gears with two rather than one sets of compounded gears in the train, I can get get down to 0.01% error or less on the common metric and BA pitches. That's 1mm out over 10 metres of thread, or 0.0001" over 1", and my Chinese leadscrew may be no more accurate than that in the first place.

Andy

[Bogstandard]

Andy,

I know you can get lathes that cut every thread under the sun (at a price), but I have always baulked at cutting BA by single pointing.

I know it can be done, but the ease of which it can be done with taps and dies makes me shy away from it, due to it having no real basis about it. It seems that each size has it's own set of rules.


Just as an aside to this topic.

I am sure that with a little work you can get a mod 1 x 127 in there.

Next year, I will be playing about with mod hobs, and if it works out OK, I won't mind cutting you some gears up to play around with, I have a full set of hobs, mod 1.25 down to mod 0.3.
But please bear in mind, it is when I get around to doing it, not when someone wants me to.

I am still trying to find some bits that I promised another member ages ago, but because I can't get to them, I have to wait until someone comes around to do it for me, and that could take days to go thru all my stash points. Two failures up to now, but they are there somewhere.

Bogs

[RichardShute]

I’ve been puzzling over the remark which Bogs made before his data got lost, and have come up with these approaches to “going in at an angle”.


Method 3.
<....> The topslide is at 30º to the work and at 60º to the cross-slide. <....>Trigonometry shows that advancing the topslide when it is set at this angle will advance the tool into the work by half the amount indicated by the topslide dial. So in this case the cross-slide and topslide can be advanced by the same amount as shown by their dials to produce the same result as Method 2 with its parallel topslide. This avoids the bother of remembering by how much the cross-slide has just been advanced and then advancing the topslide dial by half that amount. Each dial is advanced by the same amount in Method 3.

Am I on the right track. or have I got myself completely lost?

Andy

Andy,
Your trig is correct, but the scheme is that all feed is made with the topslide, other than a very fine final cleaning cut taken as a plunge striaght in. The cross slide is withdrawn after each cut and then returned to the same place on the dial each time (conveniently zero) before the next cut is taken, the feed being made with the top slide.

It is well established practice to use a compound slide at a fine angle, often 6deg as the trig works out to a convenient 1:10, for taking very fine cuts or when using a tool post grinder.

As a very near approximation to 127 you could use a ratio of 80/63 = 1.26984 which really is very close, only 16 parts in 127000 error. That may be the near miss you were referring to, it is fairly commonly used within the standard set of Myford gears I believe, 'though your gear set may well be different.


Richard

 

[BillTodd]

You beat me to it Richard

I was going to suggest 47/37. Unless you are cutting a very long thread or require extreme precision, the error is insignificant.

Bill

[andyf]

Bogs:
That's a very kind offer. When you get round to your hobbing experiments, and should they involve making some large Mod 1 gears, and if 127T happened to be a convenient size to try, I'd be very grateful. I could soon knock up a bigger banjo to accommodate it. But I know you always have a lot on your plate.

It is indeed easier to use taps and dies for BA threads because they are all so small, and I've never actually single-pointed one. While working out my gears for metric threads, I also calculated a few BA ones (see link below). At the time, I was probably trying to distract myself from doing that year's tax return    I think BA was developed as a metric thread, the logic behind it being that, as the BA numbers increase, the pitches decrease by 10%. So, 0BA is 1mm pitch and 1BA is 10% less at 0.81mm, and so on through the range, with the pitches being rounded to the nearest 0.01mm. There doesn't seem to be a corresponding logic to the diameters, though.

Richard and Bill:
43T and 37T are a handy combination which wouldn't take up much room, but I have neither. 63T and 80T are in my standard gearset, but the lathe has no quick-change gearbox, and once 63T/80T are on the banjo, there isn't much room for anything else. In any case, 63T/80T produces an error of about 0.125%, whereas most of the combinations I worked out here
< http://andysmachines.weebly.com/better-mm-threads-from-inch-leadscrew.html >
produce much less error.

Regards,
Andy
 

[RichardShute]

<....>

It is indeed easier to use taps and dies for BA threads because they are all so small, and I've never actually single-pointed one. While working out my gears for metric threads, I also calculated a few BA ones (see link below). At the time, I was probably trying to distract myself from doing that year's tax return    I think BA was developed as a metric thread, the logic behind it being that, as the BA numbers increase, the pitches decrease by 10%. So, 0BA is 1mm pitch and 1BA is 10% less at 0.81mm, and so on through the range, with the pitches being rounded to the nearest 0.01mm. There doesn't seem to be a corresponding logic to the diameters, though.

Richard and Bill:
43T and 37T are a handy combination which wouldn't take up much room, but I have neither. 63T and 80T are in my standard gearset, but the lathe has no quick-change gearbox, and once 63T/80T are on the banjo, there isn't much room for anything else. In any case, 63T/80T produces an error of about 0.125%, whereas most of the combinations I worked out here
< http://andysmachines.weebly.com/better-mm-threads-from-inch-leadscrew.html >
produce much less error.

Regards,
Andy
 

Andy,
BA threads are indeed metric and are a rare example of a system which follows a prescribed set of rules rather than just arbitrary convenient sizes. There is a description here:
http://www.enginehistory.org/british_fasteners.htm

The pitch of threads in the BA system is the 'base' dimension from which all other aspects of any given thread follow. The pitch is 0.9mm rasied to the power of thread number, hence 2BA is 0.9*0.9=0.81mm pitch etc. There is a formula quoted for the diameters in the reference, it's derived from the pitch, not vice versa.
One useful characteristic of BA threads is that the clearance drill for each thread is the tapping drill for the next but one thread in the series, the same drill is also the screw head counterbore size for the next but one down as I recall, but that warrants checking. Doesn't matter so much now, but when drills were an expensive asset, reduced tooling was handy, and is part of the reason why it was a common tendency for assemblies using BA threads to use all even or all odd numbers.  

So for example a tapping drill for 2BA is a clearance for 4BA and a counterbore size for 6BA; the same holds for 3:5:7 etc.

Just a comment, but 80/63 is 0.0125% error, not 0.125%, but it's acedemic if you can't fit other gears on the banjo at the same time. Are you aware of the little application to calculate gear ratios available here:
http://www.lathes.co.uk/threading/nthreadp.exe

It's handy for finding the best near-miss given what gears you have available.

Richard

[andyf]

  Wow, Richard - I didn't know all that BA theory.

I did use "nthreadsp" to help calculate my setups, but because it only allows for two driver and two driven gears, whereas I was using three drivers and three drivens, things got a bit complicated.  And most theoretically very accurate combinations proved physically impossible to set up. To avoid frustration I created a spreadsheet with 40T (my spindle gear) as the first driver gear and into which I could enter another two drivers and three drivens. The spreadsheet then tried all possible permutations (using 40T as the first driver) of those driver and driven gears and advised whether any of them were possible in real life. 

Andy

[BillTodd]

whereas most of the combinations I worked out...produce much less error.

Excellent:)

Given your lathe is only 300mm between centres, and even the longest of thread will only be a few tens of microns out; Why bother with a 127T gear at all?

Bill

[Bogstandard]

Thank goodness I got rid of my old lathe when I did. I used to spend hours calculating out gear ratios to give me something close to what I was after. Then overcutiing slghtly to match internal to external (if you were only cutting one).

With the way my brain is addling nowadays, I would have given up on single point threading if my lathe didn't now do it all for me.

I was very dubious when I first swapped over, as with a gear train I could always get somewhere near where I needed to be, then fiddle it afterwards, but it is now just a flick of a lever, and sometimes one or two gear swaps, and it is exactly what is needed.

As long as it is standard imperial TPI and metric pitches.




Bogs

[doubleboost]

I think i can finally put this thread    to bed
I had a visit from Rob Wilson today naturally the subject of screw cutting came up .
Rob in less than a minute showed me a very simple way of checking the compound slide angle.
The tool was simply placed on the compuund slide at 90 degrees to the lathe bed (as it would be to make a cut)
the compound slide should be at the same angle as the trailing egje of the tool.

this picture has the compound set at 30 deg as you can see the tool trailing edge is not the same as the compound

this picture has the compound set at 60 deg the tool angle and the compound are the same
The video will be edited
I have one more chuck to make i will be using the 60 deg mark
Regards
John

[Ned Ludd]

Hi John,
When you next cut a thread, will you let us know if you think it cuts better than before, then you can put the issue to bed. If after four pages you don't find it to be better, it would have been better to have kept my mouth shut or at least let my keyboard rest.   
Ned

[andyf]

Andy: whereas most of the combinations I worked out...produce much less error.

Excellent:)

Given your lathe is only 300mm between centres, and even the longest of thread will only be a few tens of microns out; Why bother with a 127T gear at all?

Bill

Good point, Bill! It would be a bit silly of me to chase after a degree of accuracy which won't, in practice, make a blind bit of difference.

Andy

[Pete.]

John thanks for the follow-up. It's always good to get a fresh set of eyes on a job.

Your video is one of the better ones and very clear and easy to follow, lots of people will gain confidence from it. I like your bronze casting ones too.