screw-cutting basics

[minerva]

As a “returnee” to our all absorbing hobby I have reached that point on the learning curve whereby many of the long forgotten skills are returning to haunt me. The mysteries of heat treatment  and single point screw-cutting  to order to mention but two .
I currently have two projects on the go, viz. - the manufacture of a George Thomas's Versatile Dividing Head (VDH) and a worm wheel controlled Rotary Table. Both projects sourced from Hemingway Kits here in the UK.
Both projects have one thing in common the cutting of bespoke threads. The main spindle assembly of the VDH for example requires the cutting of three different sized BSF threads namely:
 1. 1/16'' x24 TPI    1''x 24TPI and finally  5/8'' x 24 TPI.
Coupled with the fact that I have recently upgraded my mini lathe to a 9x20 which came fitted with an integral thread cutting gearbox I felt some practice was called for before I committed valuable project component material.
I read and re-read Martin Cleeve's excellent book “Screw cutting in the lathe” (number 3 in the workshop practice series) and managed to divine sufficient information about the various thread diameters to contemplate an attempt at the various threads involved.
I possessed no dedicated thread cutting tooling and decided after seeing this post

        :proj:http://madmodder.net/index.php/topic,3523.25.htm

to have a go at making my own, the machining was fairly straightforward and went  quite well but come the time for heat treatment I was forced to look up the procedure and was most surprised when the simple procedure of both hardening and tempering of the Silver steel blank went without a hitch and after touching up with a diamond slip stone I was ready to go.

At this juncture I will confess to an almighty 'C**K-UP – I, through lack of concentration bored the main casting of the VDH some 20thou. Oversize which has meant  a rethink on the sizing of the spindle dimensions which is no big deal as the threads involved can remain as per the origional design ( I hope) the solution requiring only a replacement spindle blank.

The dilemma now was quite straightforward do I turn the male threads first and make the nuts to fit or vice-versa ?
 My strategy was quite simply to try initially each option in turn in order to ascertain the most successful. Now the results whilst satisfactory raised more questions than answers. Making the nuts first proved to be the quickest of the options altho' the fit was somewhat loose giving doubts as to the accuracy of the minor bore dia. Divined from the formula in Martin Cleeves book. Machining the male thread first followed by the nut took much longer to achieve but a much better fit was achieved.
My question is quite simply.
What do you old hands do? Have you a rule of thumb to determine the various bores, dias etc. and what factors determine whether the  male or female threads are cut first.
Hoping this rather wordy post won't bore too much
Kindest regards
TerryT
AKA minerva




Edit : link fixed  ,Rob

[Lew_Merrick_PE]

Terry,

As a general rule (i.e. subject to exceptions) you will want to turn the female thread first and use it to gauge the male thread.  The theory here is that you can just move all your set-up aft to place the (generally shorter) female thread over the male thread rather than move things off-lateral position to fit the male thread into the female thread.  Does this make sense to you?

[BillTodd]

Hmm, interesting

I usually make the male first - it's easier to measure (just wires & a mic)

Bill

Terry,

As a general rule (i.e. subject to exceptions) you will want to turn the female thread first and use it to gauge the male thread.  The theory here is that you can just move all your set-up aft to place the (generally shorter) female thread over the male thread rather than move things off-lateral position to fit the male thread into the female thread.  Does this make sense to you?

[minerva]

well  Lew / Bill
as I suspected there may not be any laid down answer to this question but from a novice position there must be an acknowledged"way to  go" in order to put oneself in the ballpark of a reasonable result.

Bill,
I have read of the wires and Mic method but never tried it perhaps now is a good time to try it, I am going over to ARCtrade euro tomorrow so I will suss out the relevant kit
Lew
I take on board what you say and it makes sense. You make no mention of the bore sizes used.Do you divine these from data books/tables or rule of thumb ?

[ieezitin]

Assuming you are turnng between centers or using the tailstock with a live center i would make the male thread first and get the piece close who cares if its under a little, if you take it out to test on the female its a little tricky to put back while engaging it back with the lead screw. All to do then is to make the Nut and keep testing it to the male piece.

If you're not using the tailstock it dos'e not matter. keep on testing brother and have a test fest.


Anthony.

[Fergus OMore]

GHT gave a full construction article on the VDH in the Workshop Techniques book and much on the small rotary table in Model Engineers Workshop Manual. As I wrote to Minerva- I have built many of the tools  and as an amateur found little criticism of the notes. My problem was initially lack of experience- and lack of GHT tooling.

[minerva]

It goes without saying that both of GHT books are in my possession and in all honesty I find them rather confusing at times not so much from the technical content but more from the editing point of view. Given that the book was almost certainly composed from magazine articles they have a tendency to be chronologically muddled. For example the first nine pages of chapter 17 describe the initial machining operation in some detail, page ten then informs that if mounted the other way round the setting up is halved and the chances of a more accurate result increased. As to the machining of the threaded parts of the spindle and associated nuts ,his advice Nuts first but no info on bore dia which still makes it a somewhat  empirical approach. (suck it and see).
I have just returned from Arctrade Euro where I have acquired a set of thread measuring wires so I will have a look at them this evening and see what I can learn.

[Lew_Merrick_PE]

Terry,

Other than the fact that I have a (relatively complete) set of Whitworth taps & dies that dates to my days of driving MG's & Coopers, everything I "know" about Whitworth threads is what is published in Machinery's Handbook.  If you go to http://www.scribd.com/Lew%20Merrick you will find papers on design and application data for UN (60°-V), Acme, & Stub-Acme threads.  (I apologize for the mess that is now ScribD -- they were taken over by Facebook and, as soon as I can find a better spot to post, I will be closing things down there.  ScribD is nearly worthless now that the social media trash has moved in.)  The rule of thumb for 60°-V threads is that your clearance diameter for internal threading: D_i = Major_Dia - 0.75/TPI.  You would lose a bit of the minor diameter from an internal thread using that for a 55° included angle (Whitworth) nut, but that would be (about) .016 inches/inch of diameter (assuming 75% of full thread fit).

I thought the question was whether it was better to mate a (one-off, assumed) male thread to a female thread or a female thread to a male thread.  In most cases, moving the set-up clear is easier using the female thread at the gauge -- and I assumed that was the actual question.  If your threads have to repeatably match a designated fit standard, then making a Go/No-Go gauge using the overwire measurements is the way to go for dealing with the female threads.

[minerva]

Lew,
your assumptions were correct and having read your dissertations I begin to understand the problems involved .The "Rules of Thumb" that I seek seen to depend upon so many factors, % of thread engagement , angle of thread to name but two. I have quite unwittingly made two out of the three nuts which I will use to gauge the fit on the spindle
many thanks for your input
Kindest Regards
Terry T

[Lew_Merrick_PE]

Terry,

In most cases, a nominal 75% of full thread (+/-5%) is more than good enough.  The factor of safety calculated (or assumed) in most designs make "gains" from increasing percentage of full thread irrelevant.  I work in areas (aircraft, spacecraft, high-load vehicles, etc.) where I often need to know such things, but, as one friend of my likes to say, There's the few -- and then there's Lew...

The flip side of that lies in designing parts for mass production.  Reducing the percentage of full thread increases tap or die life which reduces production costs -- an attribute that concerns few here like it does when designing something that is going to be made in multi-million units on an annual basis.  I often design things that are made in 5-15 million unit annual quantities, so I think that way -- and everybody here (should) already knows that I am strange...

The history of the development of (modern) screwthreads is really quite interesting.  Ernest Lowenhertz began this process circa 1750 and laid the foundation for virtually all that followed.    The NAZI's tried to turn Lowenhertz into a Hero of Aryanism, so he has been written out of history as punishment.  Joseph Whitworth continued (and simplified) his work.  (Brian?) Sharpe carried it further (setting up the 60°-V thread) and can be argued as the creator of the system of fits we use today.  The mathematical/engineering understanding of the interactions in screwthreads did not really appear until the end of the 19th century.  We who have the answers pre-calculated and tabulated rarely understand the amount of work that developed such an otherwise simple mechanism.  Even with all these advantages, it is still somewhat hard to grasp the minutia of a well-implemented screwthread.

[minerva]

Lew
I originally started this thread in order to nail down a practical course of action when faced for the first time, with the  single point  cutting of several different sized threads somewhat larger than those normally encountered . The every day use of the ubiquitous taps and dies it must be said does not encourage one to retain knowledge of each and every size, normally the tapping size drill tends to stick in the mind and one tends to calculate this figure as  maj.dia-pitch  ( I made a conscious decision to go metric when I started back some four years ago).
I have decided to invest in a set of thread data tables as a start to my edification into the mysteries of alternate thread cutting and the application of specific threads for specific jobs, the learning curve seems to have got steeper!!
once again many thanks,
terry

[andyf]

Terry, I do it in a somewhat rough and ready manner, using the tap drill sizes shown in tables.
Simple example:
Thread required is M10 x 1mm. An M6 coarse thread has 1mm pitch, and tables show the tap drill size for it is 5mm, which is 1mm less. So to screwcut an M10 x 1 female thread, start with a hole which is 1mm less than the major diameter of 10mm, i.e a 9mm hole

More complicated example, using one of the threads you want:
Thread required is 1''x 24TPI BSF. Thread tables like this
 http://mdmetric.com/tech/thddat8.htm#bsf (BSW is at the top of the page, BSF at the bottom) disclose there is no standard BSF thread with 24 TPI. But a standard 3/16" BSW has 24 TPI. The thread form is the same for BSW and BSF. So (converting to metric) subtract the tap drill size of 3.70mm from the major diameter of 4.76mm. The difference is 1.06mm. The size required for 1" X 24 BSF will be 25.4 - 1.06mm = 24.34mm. The answer would be the same if the instructions called for 1''x 24TPI BSW.

As I said at the outset, a bit rough and ready, but it seems to work for me when making a female thread in one part and a male one on another, and neither will be required to mate with anything but each other, or another thread I've made myself.


Andy

[Lew_Merrick_PE]

( I made a conscious decision to go metric when I started back some four years ago).
I have decided to invest in a set of thread data tables as a start to my edification into the mysteries of alternate thread cutting and the application of specific threads for specific jobs, the learning curve seems to have got steeper!!

Terry (& all),

The best data set on (most) screwthread forms is (U.S. Government) FED-STD-H28.  It can be downloaded for free from http://quicksearch.dla.mil/ -- it is a bit complicated, but all the information (more than any sane person would need) is there.  It is non-classified, public domain information.

For God's sake, do not pay for thread data!  There is quite a bit posted in my "page" at ScribD (http://www.scribd.com/Lew%20Merrick) and, if you have trouble downloading it, drop me a line (tangent@olympus.net) and I will e-mail it to you.  There is a public domain booklet that was published by the U.S. Government during WWII on the mathematics of cutting screwthreads.  My copy is buried (we moved a few years back and I do not have space for my technical reference library, so there are boxes of books sitting in storage), but I will try to dig up the title of it.  It was aimed at taking farm-hands and teaching them how to cut threads on a lathe.  It was one of the texts used back when I was an apprentice...

Please remember that about 15%-20% of the work I do is seriously mission critical (i.e. the bolt circles that held the parts of the Space Shuttle's External Tank together type of stuff) and I tend to think(?) in such terms.  Nearly 50% of the work I do must meet very high (medical, military, space, etc.) standards and I tend to think(?) in such terms.  This is why I tend to restate the question in my response because I regularly see applications where the quick & common answers are unacceptable.  This does not mean that I do not know or use the quick & common answers, merely that much of the time I cannot use them because of the exceptions that exist to all rules...

[minerva]

Hi Andy,
lots of positives here.
Firstly many thanks for the link which I feel has most of the information I shall ever require. also the details of their use. your explanation has clarified the whole operation  I  have always suspected that matching male and female threads cut with the same profile must  if machined to fit  be approaching something like the definitive version should be.
Lew
I tried to download from the quicksearch site and encountered some pretty heavy security warnings ,so I have postponed for now and will try again later when I have a little more time.

Regards
Terry T

[andyf]

Just remember its limitations, Terry (my idea, not the table).

  Are you sure 1/16" x 24 is right? That will be a bit hard to do on the lathe, because a threading tool to fit down your (approx) 0.5mm pilot hole would be impractical. So would screwcutting the male thread, with its 0.5mm minor diameter, ready to snap off as soon as the tool touches it.

You might consider converting the two other threads to metric, if you're more used to that. 24TPI is close to 25.4 TPI, or 1mm pitch. 1" is only a little over 25mm, so you could substitute M25 x 1. 5/8" is only a little under 16mm, so you could substitute M16 x 1. The pilot holes would be 24mm and 15mm. 

Andy

[minerva]

Andy,
The dimension should read 1-1/16 perhaps I posted wrong.  This size I have managed to cut using the thread data obtained from Martin Cleeves book.
I take on board  your comments  regarding conversion to metric and it makes a lot of sense its just a matter of familiarity which improves with every manipulation
Regards
Terry T

[andyf]

Sorry, Terry - I see you wrote 1.1/16. Obvious, now I look at it.

Andy

[chipenter]

Personally I would make it 1 1\8 x 12 which is Myford thread , you will have to make back plates and other fixtures to mount a chuck at least , the chance of buying anything with a 1 1\16 thread is slim .

Jeff

[Lew_Merrick_PE]

I tried to download from the quicksearch site and encountered some pretty heavy security warnings ,so I have postponed for now and will try again later when I have a little more time.

As I posted earlier in the thread, ScribD was taken over by Facebook a while back.  They have made it harder to download things from ScribD and I (and many others) are looking for a true sharing website to replace this mess.  In the meantime, anybody having trouble getting the information from ScribD is welcome to e-mail me (tangent@olympus.net) for copies of the documents still posted on ScribD.

What mind?

[minerva]

Jeff,
firstly may I welcome your comments as subject of your first post on MadModder
The project in question (GHT's versatile dividing head) is as its title suggests  adaptable to most tasks asked of it and was I suspect designed to be  manufactured on a Myford for use on a Myford, its versatility depends to a great extent upon the various inner spindle additions
The threads concerned however are for the express purpose of retaining the outer spindle  within the main body casting so there is no need or reason for modification of the design at this time. The inner spindle add- ons for fitment of chucks etc.I think will ultimately depend upon the machine that the tooling will be ultimately used on.
Regards
Terry T

[Fergus OMore]

The 24TPI is simply because the number factorises with 8TPI on a Myford( and several English Lathes)

On the T and C postings, I make some comment about my GHT VDH. Mine has a solid Myford spindle but really shouuld be changed to accept Myford and ER Collets.

This is for MY way but others must find their own developments. The design allows for such changes.

[minerva]

Norman ,
I suspected as much but thus far I haven't given the latter stages of the build much thought, however having so said I use  my ER32 collets in both the lathe and the mill so an adapter to fit the VDH would fit the bill a treat.
Regards
Terry T

[Fergus OMore]

Terry,
            You are using a 'Metric' lathe although you maybe working in 'Imperial' for dials etc.
GHT used 24 TPI because many Myford etc users couldn't afford the gearbox and made do with change wheels. Having said that, most of his writings apply equally to the 920- if you watch what you are doing. There is  no real reason to stick with Imperial- if you are happier going metric/metric. If you are working from a 'non kit', you will find metric kit/metal far cheaper and more available.
There is no reason that I can think of not to metric thread on an imperial diameter shaft. After all, 24TPI is a 'Bastard' thread and so is the Myford which is Whitworth thread but should be UNF 60 degrees. Don't ask- but I fell for that one

As for this lot of nuts, please realise that all that you are doing is allowing the spindle to turn and another holding the gear wheel against the spindle-- and both have pegs and - grubs screws. I have pegs galore- dropped somewhere and I know the problem-- but can't find the things.

I have no idea where you are at but I would seriously look at several GHT 'goodies' to help.
I would have( and did have) a steel sub table with a GHT rear parting tool on it. I had a vertical slide/vice from a ? that had also been on a Myford. I have a boring head( and an aching one sometimes) but I made up a pair of inline boring bars again GHT and did the Quorn.
Finally( yea I've got more GHT) but I'd make up a mandrel handle to sort out the speed which at 130 rpm, I can't cope. You may, I can't. I fitted and inverter/now on my Myford.

Concluding, you will enjoy the re-learning experience

Cheers

Norman

[minerva]

Norman,
The kit from Hemingways is unashamedly Imperial as are most of GHT's designs, and as a 70 odd year old having grown up with old numbers they hold no particular terrors for me . I do however take on board the comment as to the availability of Imperial sized stock it is certainly becoming as rare as rocking horse dung!
I do have as part of my kit a good quality boring head with which I have had good results, but I will be making one of GHT's in line designs as his comments as to the direct measurement of dimension achieved appeals .
I have manufactured an adapter to facilitate the use of my vertical slide on the 920 , which only leaves the rear mounted tool post , Although I must say parting off is one skill that I have retained and thus far have had no real problems from the QCTP in the normal position.
I re-iterate how much I value your (and all other contributors) input . and  the thought processes  thus provoked makes membership of this forum so worthwhile.
Kindest Regards
Terry T

[udimet]

Terry,
                 Go to the VARGUS website and download the free programme VARDEX TT, it will give you all the info you could want and will also calculate the DOC for a given number of passes. Fantastic programme. Just make sure you download TT  [Thread turning] and NOT the programme TM [Thread milling].
                                                                                                  Best wishes.
                                                                                               UDIMET.