whats wrong with my thread/screwcutting

Thought I would try and replicate the thread on my toolpost lever.

here it is,

seems to be 22tpi using my handy gauge.

I used the set over method with the top slide set at 27.5 and my tool at 55.

This is the result.

It looks not bad but will only screw into the corresponding hole by about 3/4 a trun then it binds. The profile is not symmetrical , - I think due to the way it was produced. I used a mandrel handle.

Any thoughts /ideas/ tips/ pointers to a bettr technique would be good.

ps , measures spot on in diameter.

You have probably not set your top slide to teh correct angle. It should be 27.5deg from a line ar right angles to the lathe axis not 27.5 degrees from the lateh axis. On most toolposts you need to set it to read 62.5 (90-27.5) on the scale.

Jason

PS its been covered in detail hear before but teh search is pants.

I thought it should be 60 degree for uts or metric

55 degree for whitworth or bsp

I could be wrong of course

Any thoughts /ideas/ tips/ pointers to a bettr technique would be good

If you're getting weird results when you shouldn't get weird results...then use delrin.

With delrin you can cut your thread in 5 mins flat and then test it

Alternatively, use aluminium

When you get it right the thread should be a doddle to screw in, no resistance, nice and snug.

Then do a proper steel one.

55 is the included angle so you set teh topslide to half that so the tool runs down one side of the thread, look at the following pinched from a post on another forum

Hi Capnahab

It looks like a left hand thread from your photo. Were you cutting from right to left, that is toward the chuck? or from left to right?

Alan

Are you sure that the original thread is 22 tpi? I would start fault finding with that, measure over a few threads and do some calculations. 22 tpi would measure .136" across 3 threads.

Also, are you sure it's an imperial thread? This pitch is close to 1mm pitch!?

Andy.

thanks Guys , a pint to Jason so far for the correct diagnosis.

what book is that Jason ?.

clockie , it is not left hand (maybe my photography).

Only screwing on 3/4 of a turn could be caused by one or more of several factors.

First I would establish what thread the original was. If you know the OD and aproximate pitch then by looking in some thread charts you can establish if its metric or imperial, Whitworth, UNF or whatever. If you have cut a close fitting thread but the pitch is only slightly out then it wont screw on very far.

Using a micrometer and two wires that is a diameter which sits on the flank of the threads, compare the original with the one you made to see if there is any clearance at all.

From a practical point of view, if this part is only for a lever that is going be fixed permanently in place then go with what you have got and get BRUTAL! Use vice grips to install it and you wont need any threadlocking compound.

Ian P

Edited to correct poor spelling

Edited By Ian Phillips on 12/03/2012 21:04:35

I don't know what book it is I just picked it up from another forum rather than scanning one of my books and risking copyright issues.

As I said its been covered on here in quite good detail with plenty of pics and diagrams but I can't find the post.

Andy, as its 22tpi and being cut at 55deg I would asume the OP has already determined the thread is 5/16 BSF.

Ian, The second picture clearly shows that one side of the thread is being cut at 27.5 deg and the other at a much shallower angle, this is the classic sign of the topslide being angled from the wrong axis.

J

Edited By JasonB on 12/03/2012 21:09:45

Edited By JasonB on 12/03/2012 21:14:42

There are some other catches too. You say spot on diameter and are probably comparing original and copy but often people forget the crests should not be pointed and are flat or rounded depending on thread by about 1/8 depth. So the starting diameter is less than nominal. If not the nut can bind on the crests.

Then the depth you cut depends on that starting point and whether your tool is pointed, flat, or rounded. The flat and round depend on pitch so every one is different and you may be using a random ground tool. This affects the infeed.

Since the calculations still don't account for the quality of your nut it is easiest to start right, get close, then just try the nut until it fits knowing the binding is not the crests but the roots so you just need to cut a little deeper.

Hi
If the thread is the same as the job ie RH Why not set the top slide @ right angles to the job using the x slide for depth of cut . then try if its still tight At the same depth you can take say 2 thou of the the flank of the thread . Etc I good you are using the mandrel handle Is this a thread in a thread ?
Nobby

I think you would have less trouble if you set up, and go straight in,as Nobby says, rather than setting over, make absolutly sure that the tool is correctly set, take your time, and if you can find a nut that has the same thread, it will make it easy to check your progress.

You don't have to be too fancy, my first thread was a 1" whitworth one, and the tool was a bit off an old sqare file, the only problem with whitworth is the round creston the thread, quite noticable with 1", I remember at school we were taught to use a chaser to finish the thread, this is not needed with SAE, or Metric threads. Ian S C

capnahab,

There is little mention at present about checking the effective thread dia. Using three equal small wires you can mic over the original one and then repeat on your new one .

It seems from the photos that your new thread is sharper crested than the old one and that might indicate that you need to go deeper or widen the thread.

Another quick way to compare the 2 thread forms is to lay the two threads together against a bright sky. This would confirm your TPI and give some indication about the match of you thread profile.

As a rule of thumb you don't often find metric threads mixed with imperial dimensions. A check of the surrounding diameters and lengths will often confirm which you are dealing with .

Bob D

The first photo looks like a square thread to me, a white background would help a lot, if you offer your cutting tool up to the original thread does it fit well?

Andy

What make of lathe is the handle off, that would give us some idea of the thread, if it's british, theres a good chance it's a whitworth thread, otherwise, unless it's from the USA, the threads will be metric. Ian S C

For W.W. form threads I always set the top slide to 25 deg. as recommended by Geo. Thomas. This makes the trailing edge of the tool take a fine skim and results in a smoother thread. It also makes the calculation of the feed depth easier!

Capnahab, if you want a reference book with a good treatment of screwcutting I can recommend Geo. Thomas "The Model Engineers Workshop Manual".

Russell.

A new contributer to MEW issue 188 also appeared on this forum yesterday, Rick Sparber. His website has two tutorials on thead cutting. Several days of reading there actually. Help me to hit my download limit with something useful for a change.

http://rick.sparber.org/index.html

this is the first one http://rick.sparber.org/ftt.pdf

and more http://rick.sparber.org/spt.pdf

thanks for all the helpful replies.

The toolpost is Aloris . I emailed them , - thread is 3/8 , 24 tpi. So the suggestion it could be a 1mm thread is probably right .

Had a go with the thread gauge again. its not that easy to use, seems 1 mm .

Will try again.

Nice site from rick.