|Suffolk Brian||14/07/2017 15:50:39|
|15 forum posts|
I have a gunmetal casting which is to be the main part of an axle pump on a Metre Maid Loco. (0-6-0 version of Sweet Pea.) The bore of the casting is 7/16, the O/D 5/8 inches. The drawing calls for a 26TPI thread on the 5/8 OD, this to fit into an overgrown "washer", bored to suit, 1/4 thick, in this case 1&1/2 diameter. I have a pair of single point bits of 60 degree angle. I am thinking to do the pump casting thread first , then just make the "washer" to fit. I have read through Tubal Cain's Drills Taps and Dies, and although I know more than I did the book does not deal with single point thread cutting. I am mainly concerned with how deep I should cut (the thread depth) for a reasonable thread engagement, (60%?) to then give me a root diameter for the "washer".
Comments/opinions/ suggestions gratefully accepted.
8719 forum posts
This is how I tackled a similar problem recently. It might not be the best way but it worked for me. You may wish to see other answers before proceeding.
I'm metric, but bear with me:
First I cut a thread .87mm deep on the external diameter and, while the work was spinning in the lathe, I gently used a file to just take the sharp top off the thread. Then I polished it with the abrasive back of a washing-up sponge.
Second I calculated the diameter of the hole. In your case, the diameter of the hole needed for the internal thread will be 5/8" less 0.33mm (0.013" ). I make that 0.612"
Third, I made a test piece to check the calculation. I suggest you bore a 0.612" diameter hole in a short length of aluminium and then cut an internal thread 0.022" deep. Without removing it from the lathe offer up the external thread and see if it fits. It should look as if it will 'go' but you may find that you need to go slightly deeper than 0.022" before it does. Don't overdo it, the tighter the fit, the closer the engagement.
Fourth, having made sure all was well I made the real one, again by cutting the thread deeper step by step until it fitted.
The key feature of the process is that although maths is used to make a start, the actual matching of the threads was trial and error. In other words, the depth of engagement depends more on patient fitting than working to an exact specification. As you're on a lathe, you should be able to get better than 60%,
I'll be interested to see what the experts say; I still have a lot to learn. Apologies in advance for my mistakes!
(The numbers come from the metric thread specification, which is expressed in terms of H, where H is approximately Pitch X 0.87)
Edit Removed some hateful automatic smileys!
Edited By SillyOldDuffer on 14/07/2017 17:06:49
Edited By SillyOldDuffer on 14/07/2017 17:12:49
|1504 forum posts|
Tubal Cain would give you the thread depth as a fraction of the pitch (crest to crest measurement) which for 26tpi would be about 0.6, to give you a rough idea. Dave's method of "sneaking up on it" is one which many of us would follow.
|Suffolk Brian||18/07/2017 12:24:22|
|15 forum posts|
Grateful to Dave & Robbo - very helpful. Thank you both.
|Martin of Wick||08/11/2017 22:50:34|
|249 forum posts|
Sorry to resurrect an old thread, but the original post noted the general lack of guidance or principles in the literature for screw cutting with a single point tool. ME handbook is a good start, but is hardly a practitioners guide.
If the diagram in 4.1 in the ME handbook for BSW is to be believed, then apparently there is no thread flank clearance for BSW threads - can this be correct? The thread engagement is not stated - do we assume 70% as drawn?
I admit I have no experience of cutting threads on a lathe but am planning the process for cutting an M7 backplate. I am more than happy if anyone would correct the approach I plan to use in the worked example below.
So we assume a perfectly ground single point tool ending in a sharp V and imperial dimensions.
As I am a beginner, for the avoidance of error, the tool will be used in a way to initially cut the full thread height given as 0.96p, not the truncated form. The threads will truncated in a second operation using the dimensions given in 4.1.
Therefore, if cutting an external thread of 1.125 inch nominal for 12tpi (0.08333 pitch), I would start with material that is 1.1516 diameter ie. truncated form plus 2x 0.16p or 1.125 + .0266. Then advance the cutting tool 0.0799 (based on 0.96p from 4.1)
The threads should come to a nice sharp point at the end of thread cutting and be a visual confirmation that the cut had been completed.
I would then truncate the new thread by advancing a cutting tool 0.0133 in. (0.16p) to generate the truncated form of the major diameter, restoring the external diameter to 1.125 in.
If I was cutting the internal thread, I would bore out to the untruncated minor diameter, calculate from the untruncated major diameter (above) minus 2 times the untunctated full thread height.
In this example it would be 1.1516 - 0.1598 or 1.1516- 2(.96p) that is equal to 0.9918 in.
The threads would be cut by advancing the tool by the same 0.96p or 0.0799 in. followed by a cut with the boring bar of 0.0133 in. to generate the truncated form of the internal thread.
I am relying on the fact that the nose threads on my clapped out M7 are so worn that additional thread clearance cuts will not be required, but if cutting a matching set of threads I would probably go in an extra thou or two for flank clearance -
hoping that somebody may know if there is a standard for BSW flank clearance?
Also recognise that as an alternative, you could start the internal threading by boring to the minor core diameter of the truncated form and then making a threading cut by advancing the tool 0.8p or 0.0666 in. and avoid the second operation to truncate the full height threads. But I quite like the idea that at the end of the thread cutting process you have a sort of visual check - if the flanks come to a sharp point on the final cut, then possibly you can be reasonably confident all has gone well - at least until you try to screw the two parts together and learn something else!
Of course, in trying to establish a set of repeatable practical steps, I may have completely misunderstood the approach required and got it all completely wrong- don't hold back you will be saving me from scrapping a backplate!
Also happy to hear other methods from anyone experienced in screwcutting or whether there are more reliable or simpler methodologies for use when form tools or taps are not available.
6445 forum posts
Doh, double post. See the next one.
Edited By Hopper on 09/11/2017 02:22:59
6445 forum posts
Edited By Hopper on 09/11/2017 02:27:55
|John Olsen||09/11/2017 03:09:19|
|1253 forum posts|
When I wanted to screwcut an internal thread to screw onto the nose of my ML7, I first made an external thread, test peice which I was able to fit to one of the chucks. (eg the test piece was held in the three jaw, so I was able to test the fit in the four jaw or the faceplate.) Once I was happy with the fit, I was then able to mount the actual job, a backplate for an ER32 collett chuck, and do the internal screwcutting. The test piece is then used to check the fit. It can take a surprising number of passes to work out all the spring,. The test piece might for instance go in a turn or so then tighten up. So you keep taking tiny scrapes off until the test piece goes in freely all the way. Once that is done, you can screw the backing plate onto the mandrel and turn the face true for the chuck to mount.
It might seem a bit of extra work making the test piece, but then, you will have one for next time you need it.
|Martin of Wick||09/11/2017 11:58:42|
|249 forum posts|
Thanks for that chaps, that's 2 votes for the 'cut and try' method then! - no harm in that as it is what most of us do most of the time! I fully expect to spend a lot of time cutting practice and test pieces. Appreciate the point regarding clearances on thread diameter on lathe cut threads but wondered if there were any actual published data or whether this is simply a mater of practicality?
What I was trying to get at with the post in relation to the original question was that the diagram of whitworth thread dimensions that you see endlessly regurgitated throughout the literature and now the WWW, can only precisely work if you use a form cutter (assuming you wish to get to the dimensions set out in the standard).
Even the very excellent Sparey (I have a 1948 first edition!) is as guilty as all the others in not pointing out that if you use a single point cutter at a cut depth of 0.64p from the major diameter, you will not achieve the correct Whitworth dimension of thread width and nose radius. A similar issue would apply to all types of threads due to the limitations of the single point cutter. If using a form tool, thread chaser or die head etc, then there is no issue with the dimensions as quoted.
Does this matter? - probably not in 90% of cases, as I get the impression that threading is not regarded as precision process - as long as the threaded part more or less fits and doesn't actually drop off when turned upside down seems to be the acceptable level of tolerance (although there are general guidelines and suggestions on engagement and minor diameter tolerance etc for various threads).
However, I was hoping to confirm a rigorous process from a known point of precision and then let natural expedience, practical bodgery and lack of competence determine where I end up - and when it ends in tears, at least I know that I started at the right point even if I got lost along the way!
Thanks to all.
6445 forum posts
Nobody much in the real world works to that drawing. It's simply not practical nor needed.
If you want to get a bit closer to it, instead of the radius tops on the male thread, you leave the OD that few thou undersize to allow clearance, then add a radius to the tip of the thread with a file at the end of the job. Then you can rub your screwcutting tool bit on the oil stone to put a small radius on the tip to give an approximation of the radius at the root of the thread.
If you really want to get fanatical about it, you can use those carbide insert screwcutting tools that have the radius on the tip and a radius on the crest of a form ground tooth. But then you need one tool bit for each pitch of thread, so not real practical for the home workshop.
Personally, I cheat and finish many of my screwcut threads off with a threadfile. Quick and dirty and works fine.
People have been screwcutting by the basic method described by Sparey et al for several hundred years now so it seems to work.
If you read through Tubal Cain's stuff on threads, he says you can go down to 65 per cent thread depth engagement without losing significant strength so there is no great advantage to sticking to the perfect full form thread for most ordinary work. As always, its a compromise.
PS, another way to get the perfect thread form is to use a tap of the right TPI held in the toolpost as a multi-point form tool to do your screwcutting with. Or one die segment out of a Coventry die head. I tried it once with a tap and it worked well, using a taper tap it cut very nicely over a number of teeth of increasing depth. But it was super fiddly to hold the tap in just the right position and clear the job etc. Would work better to cut a section out of one flute of the tap and silver braze it to a square bar holder like a carbide brazed tool.
Edited By Hopper on 10/11/2017 12:35:09
Edited By Hopper on 10/11/2017 12:41:02
|Andrew Johnston||10/11/2017 12:52:36|
6609 forum posts
Ah well, I must be a fanatic then. For specials I grind up a HSS toolbit, but for general screwcutting I use full form inserts. I tend to cut the same threads again and again, so I only buy inserts as needed, and they're not overly expensive.
6445 forum posts
Well, one could probably say all fanatics use form tools but not all form tool users are fanatics.
Please login to post a reply.
Want the latest issue of Model Engineer or Model Engineers' Workshop? Use our magazine locator links to find your nearest stockist!
You can contact us by phone, mail or email about the magazines including becoming a contributor, submitting reader's letters or making queries about articles. You can also get in touch about this website, advertising or other general issues.
Click THIS LINK for full contact details.
For subscription issues please see THIS LINK.