Screwcutting

[Allen Smithee]

So this afternoon I got away from work early, and had nothing on SWMBO's "little job list". I'd also received the final bits I needed in the post, so I thought I'd quickly knock off a job I've been failing to get around to for a while - knocking up some drawbars so I can use MT2 collets in my Myford's spindle and tailstock. Simple enough, although slightly complicated by the fact that I'd decided to fit them with "machine wheel" handles (the cheap 80mm chromed ones in ArcEuro's machine spares section) and a brass "thumping button" to tap with a hammer to release the collets.

The design is simple enough - some 10mm steel bar with a 3/8Whit thread on one end and an M10 thread on the other. I started by making a nylon bush that I could slip over the bar and into the spindle bore so that I could run a ~300mm length of bar through a collet at 2,000rpm without the free end whirling around. I then skimmed the end of the bar down to 3/8", turned a 4mm wide rebate where I wanted the end of the thread, dug out a 55deg tool and set up to screwcut the 16tpi thread at my lowest possible speed of 25rpm (because it's over 25 years since I last did this, and I'm chicken. Other than stupidly trying to do this with a carbide insert tool (which hung on gamely for the first couple of passes, but then cracked so I switched to a toolsteel one) the exercise went well enough with some tapping fluid, and the resulting thread just needed a light clean-up with the only 3/8W die I have - a "hexnut" die. I was on a roll. If anyone's interested I used the "plunge in square" technique rather than the "set the compound over 22.5 degrees and cut one side only" one, and it worked fine.

So I then swapped the bar around to cut the M10 thread on the other end, and that's when the fun started. I can't screwcut metric threads without fitting the metric conversion set, which takes longer than I wanted to spend on it. So I skimmed the bar down to 9.8mm just to ensure it was round, turned a 10 degree taper on the first 6mm or so to allow an easy start and placed my M10 die in a die holder on the end of the work, pushed square with the tailstock. Using a manual handle on the spindle, for the first couple of turns (on the taper) it seemed to cut well enough. But then no matter how much I tightened the collet it would only cut about an eighth of a turn before grabbing and twisting in the collet. After half an hour I had about six turns of thread and it was getting very tedious so I took the job out of the lathe and grabbed it in the vice (grooved brass soft jaws). This was a little better, but again I could only cut about an eighth of a turn before the bar rotated in the vice. After another 30-40 mins I had about 20mm of thread, but I need about 40mm and it's getting seriously tedious! I've considered cross-drilling the bar and putting a small bolt or pin through it to stop it rotating, but I'd really rather not.

Two thoughts occur to me - one is that my die isn't a spilt die, so I'm trying to cut what is a fairly deep thread in one go. Another is that possibly I need to anneal the bar a bit. But when I cut it to length it hacksawed and filed easily enough, so that would surprise me.

This is such a simple, mickey-mouse job. I don't remember it being this hard! What am I doing wrong?   

AS

[philjoe5]

Allen,
I just went through the same exercise cutting 1/4" - 20 threads on a piece of 1144 steel held in an ER-32 collet chuck on my lathe.  I used a tailstock dieholder to start the threads but after a few turns the work turned in the collet.  I had turned the work down to 0.245" and thought I had allowed enough, but eventually I found that by turning the work to 0.240" I was rewarded with some nice threads which are not at all sloppy in the tapped piece they screw into.  So I had a 4% reduction in diameter compared to your 2%.  I'd turn the work to 9.6 mm

Cheers,
Phil

[mklotz]

Genau!  Phil's nailed it.  I went through the same exercise in the past.  After measuring a bunch of commercial bolts (no doubt roll-formed so on the large side), I deduced that a -5% or so allotment when die-cutting threads is perfectly acceptable.

[kvom]

Single-point is your friend, even if you only cut partial depth.

[Art K]

Allen,
I'm with Tim, Just don't do something stupid like putting a piece of fuel hose on your crankshaft throw to "protect" it from the lathe chuck jaw's
Art

[Jo]

So I then swapped the bar around to cut the M10 thread on the other end, and that's when the fun started. I can't screwcut metric threads without fitting the metric conversion set, which takes longer than I wanted to spend on it.

Yes it does put you off having to assemble the set. But once assembled it is easiest to leave it assembled, with a note on it for what thread it is set up for and sitting on its box ready for next time.

Yes I too should remember to turn down the spindle speed when I cut threads but its so easy to forget

Jo

[tangler]

AS,

I find die cutting any thread above 1/4" / 6mm in steel a pain so I do tend to single point cut most threads and just finish with a die.  Which is where this table comes in handy:



Just slip the 24 tooth gear below the tumbler gears off and replace with either a 33 or 34 (available from Myford).  Pitch accuracy is good enough for almost all purposes.

HTH

Rod

[jadge]

It shouldn't be a problem cutting an M10 thread on a 10mm diameter bar with a die. If the die is not split are you sure it actually a die, and not intended for just cleaning up existing threads? It would also be interesting to know the provenance of the die.

Andrew

[Don1966]

AS,

I find die cutting any thread above 1/4" / 6mm in steel a pain so I do tend to single point cut most threads and just finish with a die.  Which is where this table comes in handy:



Just slip the 24 tooth gear below the tumbler gears off and replace with either a 33 or 34 (available from Myford).  Pitch accuracy is good enough for almost all purposes.

HTH

Rod

Rod there is an error on the M fine threads the second calculation with 12 using the 33 stud gear and 20 box gear. It should be 1.746 value.
For any one interested I have attached a Myford gear box calculator. It will calculate TPI and Metric threads along with gear box ratios up to three gears sets along with the Norton QC gear box.

[Jasonb]

I'm with Andrew the dienut is the most likely cause of the problem.

You could lock a couple of 3/8" whit nuts on the other end and hold by those which should help stop the bar rotating

[tangler]

Rod there is an error on the M fine threads the second calculation with 12 using the 33 stud gear and 20 box gear. It should be 1.746 value.
For any one interested I have attached a Myford gear box calculator. It will calculate TPI and Metric threads along with gear box ratios up to three gears sets along with the Norton QC gear box.

Quite right Don, there is a typo in the line - it should be 33, 28 like all the other 1.25 pitch threads.

many thanks,

Rod

[Allen Smithee]

Many thanks peeps. Years ago I used to reduce the diameter by ~5% to make the die run easier, but I'd assumed I was bodging it and I'd sort of decided to stop bodging, so this time I didn't. It's reassuring to learn that people I respect don't consider it a bodge!

The die (not a die-nut - that was the whit one) is one from a cheap set I bought somewhere over the years, so it could be that.

I'd heard about the "using a 33/34 tooth gear to cut metric threads" but never found the details, so many thanks for that info Rod. I think my first action on Tuesday morning will be to place another £20 order with RDG!

I think I'll put the bar back into the collet, get it as centred as I can and skim it down a bit.

Thanks again - just the sort of thing that makes this place so good!

AS

[PStechPaul]

The 1mm thread uses 34/36 or 17/18, and if you multiply that by 25.4 (mm/inch) you get 23.989, or just 0.04% error, or less than 0.001" in 2" (or 26mm) of thread. I made a chart of metric pitch equivalents for my HF 9x20 lathe, although it's probably better to use the 100/127 tooth gear and do it properly:
 

I didn't try a 36/34 combination. It's a spreadsheet so maybe I'll try that.
 
Here it is:

[jadge]

The die (not a die-nut - that was the whit one) is one from a cheap set I bought somewhere over the years, so it could be that.

Very possibly; many years ago I bought a cheap metric tap and die set. It was absolutely hopeless so it went in the bin. More recently I had a need to cut 5/8" UNF threads using cheap taps and dies bought 15 years ago from the well known ME supplier in the UK. Again, absolutely hopeless. Two parts made with the taps and die were such a tight fit that they had the gall to gall. No way I could get them apart so had to scrap the parts. I ended up buying a proper HSS tap from a commercial supplier and screwcut the external threads. Here is what happens to cheap tooling, it gets Widlarised before being binned:



There are two rules for buying cutting tools:

1) Never buy cheap cutting tools - they're the source of much grief

2) See Rule 1

As for reducing the diameter before threading I've never had a problem not doing so. On the other hand, for internal threads, I select a tapping drill to give 60-70% thread engagement.

Andrew

[Allen Smithee]

Again, thanks for all the above advice. I've just ordered the 33 and 34 tooth gears.

Having stared at the geartrain for a few minutes this morning I'm guessing I remove the outer gear on the shaft below the tumblers (taking care not to lose the woodruff key!), slacken the banjo bolt, fit the selected alternative gear and then reposition the banjo to re-mesh the new gear, retighten the banjo bolt and Bob's my ankle. Is this correct?

How critical is the meshing - do I just push it into contact while retightening the bolt or should I be feeding a riszla paper in there to set clearance or something?

Finally - am I right in believing that I can't use the thread position indicator when running the 33 or 34 tooth gears unless U do the trick about setting a saddle stop position and re-engaging the half nuts with the saddle held against the stop or similar?

AS

[Firebird]

Hi

I have managed to do a bit of screw cutting on my Myford and have used the set over method as described in the Myford users book.

What is the "plunge in square" method 

Cheers

Rich

[jadge]

What is the "plunge in square" method 

Cheers

Rich

Presumably the toolbit is plunged straight in for a cut, rather than at an angle as set on the top slide? All my screwcutting has been done using this method, although being idle I have moved on to using carbide threading  inserts, which are happy cutting on both flanks, rather than grinding my own toolbits. It has the advantage that the depth of the thread is directly related to the numbers on the cross slide dial rather than having to work out the feed on the top slide via trigonometry. 

Andrew

[Allen Smithee]

What is the "plunge in square" method 

Set the compound at zero degrees and just advance the cutting tool straight into the job. It means that the tool is cutting on both flanks, so you need to make sure their both sharp, but if your setup will do it without chattering it's simpler and easier to control the thread depth.

When I first started machining stuff I always cut threads this way because no one told me to do otherwise, and I've generally found that the set-over method is only really necessary for "difficult" materials.

YMMV,

AS

[sshire]

The toolbit is, in fact, a form tool, so the plunge-in straight method works just fine as with any form tool cut. The insert on the Aloris threading tool that I use is properly ground on both sides so it works quite well.
Some have suggested using the 29 (29.5, 30°) set over for most of the cut and the plunge-in straight for the last few passes. With Aluminum, brass and 12L14, I've had nice results plunging in. Haven't tried it with "difficult" materials.
I also paint the part with Dykem. When the Dykem is completely gone, I begin measuring.

[Stuart]

Allen

You may or may not need to slot the banjo a wee bit to get the new gear on the output shaft of the tumbler , I had to

As to mesh I am not saying this but I set it somewhere near and adjust but sound whilst it's running , note I did not say that at all

They need to be a bit sloppy printer paper is about right  slacker than tighter is the way to go

For lube I use the wd40 brand of chain lube you don't need much as it's very sticky , but you do need to clean off all the oil that's on them first
Just remember to keep the half nuts closed ,best to use a spindle handle for short threads, or make the Grayham Meek dog clutch drawings can be found on the web my set is for the big bore lathe do will not fit yours then you leave the half but closed and set the stops and the ad like a hardinge

Stuart

Stuart

[jadge]

I've cut threads in 303 stainless steel using the straight in technique with no problem. Although the thread was only 32tpi. Oddly these particular bits of 303 were a total pig to get a decent finish on, unlike most other 303 I have used. Despite that, the screwcutting was fine. To get a decent turned finish I was down at 260rpm with coolant, but I did the screwcutting dry at 370rpm.

Andrew

[Ian S C]

I use straight in thread cutting for finer threads, but after they get to around 10/12 tpi, I prefer to go in at the angle, reducing the load on the tool as it's cutting on a wide area by the time it's getting near full depth.  My first threads were cut using a tool ground up from an old 3/8" square file, working in back gear over short distances, with fine cuts the tool does not get hot, so it holds it's edge.
Ian S C

[Dan Rowe]

Finally - am I right in believing that I can't use the thread position indicator when running the 33 or 34 tooth gears unless U do the trick about setting a saddle stop position and re-engaging the half nuts with the saddle held against the stop or similar?

The usual advice for strange gear combinations not covered in the lathe thread chart like cutting metric with an inch lead screw is do not disengage the lead screw. This is always how I did it when I needed to cut  the wrong way on the ships I worked. I usually had a lathe with a foot brake so with a slow speed the brake can be stomped the tool backed out and reverse the machine. I could reverse with an apron control so it was a simple operation.

You should read this post to review another very workable method:
http://www.modelenginemaker.com/index.php/topic,5237.0.html

Cheers Dan

[Allen Smithee]

You should read this post to review another very workable method:
http://www.modelenginemaker.com/index.php/topic,5237.0.html

That seems to be similar to the saddle-stop method (unless I'm missing something). The method as I understand it is:

1. Put a saddle-stop on the RIGHT of the saddle in a position that has the tool clear of the front of the work.

2. Start the lathe, with the saddle held against the stop wait for the thread position indicator to get to the desired number (say "1") and then close the halfnuts

3. Make your cutting pass down the work and then disengage the half-nuts at the run-out groove or shoulder

4. Back out the tool and wind the saddle back to the saddle-stop, then set the tool (croos-slipe or compound) for the depth of the next cut.

5. Start the lathe, with the saddle held against the stop wait for the thread position indicator to get to the right number (say "1") and then close the halfnuts again.

..and so on. So essentially you engage with the saddle held against the stop, then disengage and wind the saddle back to the stop, wait for the number to come around and re-engage. Simpler and no need for continually stopping/starting the motor.

If I have it right this method ensures that the halfnuts will always close "in synch" on the same place on the leadscrew. I've evn heard of very confident people who dispense with the saddle stop and just zero the saddle DRO for the synch point.

Do I have this right? Would it work?

AS

[Firebird]

Hi

Thanks  I'll try that method next time I do some screw cutting

Cheers

Rich

[Dan Rowe]

Not to be argumentive but your method makes the unstated asumption that for a  fixed position of the carrage that there is a SINGLE position that the "1" will come up for any rotation of the spindle and gear train

This might be the case but I certanly would check it to make sure that the "1" does not come up twice in a single spindle revolution. To check simply use a marker on the stock on the top and a mark on the lead screw gear on the top and make sure that for every instance of "1" both marks are still on the top.

The method in the linked thread is a positive way to make sure all the gears and the lead screw are in the exact same relationship similar to the advice to never disengage the lead screw.

Just my 2 cents worth....

Cheers Dan

[Allen Smithee]

Not to be argumentive but your method makes the unstated asumption that for a  fixed position of the carrage that there is a SINGLE position that the "1" will come up for any rotation of the spindle and gear train

Good point, and I will check. But as far as I can see the method described by Marv makes the same assumption!

AS

[Dan Rowe]

No the method posted by Marv is a method to engage the lead screw and the gear train EXACTLY the same way.

You turn off the lathe at the same time or slightly before you disengage the half nuts. Then you move the tool bit out and reverse the lathe and engage the lead screw when the "1" comes back. The lathe will not have turned much in slow speed as this is done so you can be sure that the lead screw and the gear train is in the same orentation as before which is similar to not disengaging the lead screw.

If done correctly the "1" will only turn a few degrees and reversing will line it back up to the very same spot.

Cheers Dan

[mklotz]

Thanks, Dan.  I couldn't have explained it more clearly myself.

[jadge]

Well I was going to say I didn't quite get the method, but following the explanation from Dan it now makes sense. Not that it helps me as I don't have a thread indicator dial on my lathe. 

I took it off to fit an Ainjest high speed threading unit in its place. That allows me to screwcut any integer, half or quarter tpi thread without the need to bother with a thread indicator dial. The trip mechanism is pretty consistent; I set clearance to a shoulder, internal or external, at 3 thou. The only thread listed on my gearbox that it cannot cut is 2-7/8 tpi.  But that wasn't listed on the original thread indicator dial either so I'm not sure how you were supposed to cut it?

Of course the Ainjest unit fails when needing to cut a metric thread on my imperial lathe. There are add-ons (the Metradial) that enable this, but they are an approximation, and have a cumulative error. So the more passes taken to cut a thread the worse the alignment gets. As and when I next need to cut a metric thread I'll use a bed stop on the right hand side of the saddle to return same to the original starting position. And then stick an encoder disc on the main spindle so I can count the number of turns the spindle made during the cut and then reverse it, and by definition the leadscrew, to their original starting positions.

Andrew

[PStechPaul]

One of the main reasons for screwcutting (chasing) on an angle is that the axial cutting forces ensure that the backlash in the carriage and leadscrew is minimized. With straight-in cutting, the tool can move along the axis of the workpiece by the amount of backlash and a "drunken thread" could be produced. When I cut square threads, I had to go straight in (as with a parting tool) but I used a tool a bit narrower than the thread width and made multiple cuts with the half-nut pressed either way against the screw.
 
Also, cutting on only one edge of the tool produces a single flat ribbon, whereas cutting on two edges makes a V-shaped swarf that does not coil very well and could tend to wedge into the work at the cutting edge. The 29.5 degrees for angular cutting allows the back edge of the cutter to cut slightly along the opposite facet of the thread to clean it up and avoid the steps that occur at angles greater than 30 degrees.

[Allen Smithee]

I was wrong!!!

Apologies - having checked this morning I found that the thread position indicator is geared 16:1 to the leadscrew (ie one complete turn of the indicator is 16 turns on the leadscrew and/or 2" of saddle travel).

So this method won't work at all. There was a method that worked something like this, but I can't remember waht it was, so I'll stick to the one Marv described unless/until I can remember or work out what the method was!

Apologies for post twaddle (again) 

AS

[mklotz]

Allen

Have a look at this
,,,

That's the video I mentioned suffering through in my original post.  To paraphrase A. Lincoln,

He can compress the most words into the smallest ideas better than any man I ever met.

[Stuart]

Sigh

I was only trying to help but if the video offend I will delete the post

Stuart

[Edward]

I rather like Toms videos. Each to his own I guess but I have learned a lot from watching him.

[mcostello]

I am an avid Tom supporter and watch all His videos. I just would not want Him to be the one to yell "fire" in a crowed room.

[Tennessee Whiskey]

Stuart, I think Marv was basically saying that the content is good,  however,  if you removed all of Tom's uhs, ahs and ums, and BS,  the video would have been a whole lot shorter,   whilst still getting the point across.  If you don't mind the time involved,  Tom's videos are pretty damn good. His series on cutting short  machine tapers is excellent.  When I've read every post on the forum up to date,  I'll watch YouTube videos on processes that I'm curious as to how others perform. I've concluded that there is a group of YouTube machinist that do pretty much as we do here on the forum, they just do it on YouTube. They swap laughs, tools,  procedures,  and ideas. Stuart,  carry on.

Cletus

[Stuart]

Thanks Guys

Stuart

[Allen Smithee]

So this afternoon I got an hour or so of swarf-making time again. I put the bar back in the collet and skimmed it to 9.5mm, and then the die cut the remaining thread like a hot knife though politicians. I'm really amazed how much difference the odd 0.2mm of dia made.

I then turned my attention to the handle - this is made from the cheap 80mm chromed machine wheel form ArcEuro. It's cast iron and comes with an 8mm bore so I drilled it out to 8.5mm and introduced the tap from the same set the die came from - leaving it in the lathe with the tap held in the tailstock to start the thread square and using tapping oil. For the first 10mm or so it went in OK, but then it got progressively stiffer. I was puzzled by this because it was a 2nd tap rather than a taper tap and the full dia was well inside the material before it got stiff. By the time I got to 15mm deep I was getting really concerned that the tap would snap, so I withdrew the tap and looked at it. I'm not sure what I was looking for, but the leading edges of the teeth didn't shout "sharp" at me.

So I dug around in my old bits draws and found the taps I still had from 30-ish years ago. You may remember these are amongst the things I found had become immersed in water for several years due to an unseen roof leak in the garage. I'd found my old taps and dies were a bit rusty, and although they cleaned up quite nicely with a brass brush in a dremel I'd put them to one side as possible future boring bar material. Amongst these was an M10 2nd tap which looked like toolsteel rather than the carbon steel one I'd been using, so I gave it a try. It cut through the cast iron so easily I was initially convinced the thread had stripped! It literally took me about a minute to complete the thread through the 40mm deep hole. I then started on turning the brass thrust washer/spacer to go between the handle and the spindle, and the brass hammer button to go on the end of the stud, but ran out of time.

So my conclusion is simply that the tap & die set I was using should be regarded as a "for brass and aluminium only" set, at least in the bigger sizes (I've used the M3, M4 and M6 taps & dies on steel successfully before now) and I'll have to buy some decent ones!

AS

[Lew Hartswick]

When any video starts out with some jerk blathering away and moving like  rock star on stage I immediately hit the X and go on to the next post. :-(
   ...lew...

[Stuart]

Point taken there Lew

Post deleted

Not worth the time and trouble to try and help

I will just go to a lurker

Stuart

[Allen Smithee]

Don't let any of the old farts trouble you, Stuart. The personality defects are theirs, not yours.

I found the video interesting, and would ask you to reinstate it.

AS

[Tennessee Whiskey]

Stuart, I agree with AS, that's why they have the X button.  Some like action films, some like comedies. I can make the choice myself of which I prefer.  Repost it.

Eric

[steamer]

Don't let any of the old farts trouble you, Stuart. The personality defects are theirs, not yours.

I found the video interesting, and would ask you to reinstate it.

AS

Ditto!!!!!!

[mklotz]

I've got a four hour Andy Warhol epic of paint drying if you lot want to really live it up.  The 30 minute conclusion is shot in slow motion for extra tension and dramatic intensity.  It will leave you with a whole new appreciation of evaporation.

Or maybe you would prefer his "Empire"...

<a href="https://www.youtube.com/watch?v=-sSsWj2HWk0" target="_blank">http://www.youtube.com/watch?v=-sSsWj2HWk0</a>

[Allen Smithee]

I've got a four hour Andy Warhol epic of paint drying if you lot want to really live it up.  The 30 minute conclusion is shot in slow motion for extra tension and dramatic intensity.  It will leave you with a whole new appreciation of evaporation.

Or maybe you would prefer his "Empire"...

<a href="https://www.youtube.com/watch?v=-sSsWj2HWk0" target="_blank">http://www.youtube.com/watch?v=-sSsWj2HWk0</a>

Thanks Marv - I'll add that to my spreadsheet

AS

[steamer]

When any video starts out with some jerk blathering away and moving like  rock star on stage I immediately hit the X and go on to the next post. :-(
   ...lew...

Wow!    I guess we'll have to give you a full refund!    .....oh wait.

[Jo]

So Pete am I allowed to mention that you seem to be enjoying thread cutting so much that you are looking to buy yourself a little something to practise more thread cutting on 

It will be a good one as you have to cut both the internal and external threads and there will be no taps or dies involved  . They are nice big diameter threads but you are allowed to cheat I don't think Kirt updated those threads on the drawings to metric and they are still an inferior pitch 

Jo

[Jasonb]

Or you could build my 5cc glow/diesel for free, well a few quid for a lump of ali so not a lot lost if things go pear shaped. although 1mm pitch you could happily do then 26 or 24tpi if you prefer to work in old money.

I know a lady who may let you look at her plans if you ask nicely

[Allen Smithee]

A gentleman would never reveal whether he's had an intimate rummage through a lady's drawings.

But yes, I could be interested Jason - I'd probably put an RC carb on it though. And probably modify the front end to carry the crankshaft in ballraces.

Jo - wait until the deal is concluded and then we can reveal all...

AS

[PStechPaul]

One could say that a totally different form of "thread cutting" was exhibited in this thread as applied to a post.