thread on outside threading ...

[wquiles]

I am having some difficulties on outside threading that I wanted to share with the experts here - hopefully you guys can help me correct what I am doing wrong.

What I am trying to do is a project for a forum member in which I am basically duplicating this escrew:

The problem that I have is that when I am done, the peaks of the newly formed threads have been basically "squeezed" out and I end up with a larger OD than the one I started with. I don't know if this is due to the straight cutting of the threads (no 29.5Deg) or because of the actual thread cutting tip (60Deg), or what. I will post photos here in the hopes of trying to learn more


So I first get the OD to the same outer thread dia (major diameter):

I then make the edge beveled:

While I am threading, I am using a trick I learned here using super magnets holding a dial indicator. I have the parts from Modamag to get my Shumatech going, but this works in the meantime:

Then I cut and bevel the grove where the threading will end - this also marks the size of the heatsink:

So I am now ready to cut threads - I even have the thread table indicating the total cut depth and how much to cut at each pass (I am doing 20TPI):

So I set up the cutter for the first pass:

here is after the first pass:

second pass:

third pass:

fourth pass:

fifth pass:

sixth pass:

Even though I am suposedly done, when I try the head on the heatsink, it is still too big:

and when I measure the OD, I can see why - I am now at 0.822" !!!:

So I used a diamond file to cut the "tops" of the threads, and did another pass on the threads to clean-up:

and now it fits, although now the fit is a little bit too loose for my taste (looser than the actual escrew):

So what do you guys think is the problem?

Will

 

[wquiles]

RESERVED

 

[Mirage_Man]

Will, as you know I advocate the use of the 29.5* method over straight infeed. However I would suggest hitting the peaks of the threads lightly with a file with plenty of WD-40 or whatever you use on the next to last pass. Then make your final pass of about .001-.002". Then spin the spindle up to 1000 or so RPM and hit it with a Red and then Gray Scotchbrite again soaked with WD-40.

 

[precisionworks]

29.5° is also what I use, as it's easy to set the compound to that angle. Then a fishtail gage is used to set the tool at 90° to the part being threaded.

For almost all lathe ops, carbide tooling is used. The exception is single point threading in aluminum. A carbide insert is never as sharp as a freshly ground HSS tool, and carbide tends to tear most alloys including 6061. It's easy to get the correct point angle, again using the fishtail gage.

I also use mag-back bases for threading. When using 29.5° compound, you make a pass & retract the tool, go back to the start position, advance the tool back to where it last was, zero out the dial, and dial in more depth. Much faster to do it than to write about it:thinking:

 

[VanIsleDSM]

I think that threading table is wrong, it doesn't tell you to go deep enough.

This one is much more in depth, in 2 ways

http://www.walteranderson.us/hobbies/metalworking/microlathe/frog/threading.html

0.28" for a 24TPI thread for example, is wrong, the actual depth is 0.316", which means if you were using 29.5 you'd actually want to go in 0.365". The threading and the depth seems easy now.. it's all about getting the diameters right. I'm debating on just buying some full form 24 TPI inserts or something.. yet I have nothing to insert them into yet.. would be nice to have the bit knock down crest of the thread once you're deep enough.

 

[gadget_lover]

At first I thought you were moving the compound to the left, then I noticed the dial indicator in the back.

That's 60 degree sharp ended tool, right?

What I can not tell from your pictures is wether the tip of the tool is centered at the right height. That makes a difference. It also looks like the threads are different widths in one picture ( mg_5912.jpg ). Other than bumping the compound crank or loose gears, I'm not sure how that happens.

Is there a radius on the tip of your tool , or is it sharp?

Daniel

 

[PhotonFanatic]

Hi Will,

If you are going to cut threads, you really should learn the proper technique, especially for the coarser pitches, i.e., 4 to 24 tpi. That means using your compound slide set to 29-30 degrees to the axis of the workpiece. Doing so will place less stress on your cutting bit, resulting in cleaner threads. Sure, for fine threads, where you really aren't removing much material, feeding your cutting bit straight into the piece will be OK, but I suggest you adopt one technique and stick to it. The odds of making a mistake drop significantly once you have the routine down.

So if you are going to make threads using the 30-degree compound slide setting, here is an explanation as to how and why to determine the depth of cut:

The Theoretical Single Depth of Thread for either the Unified or ISO Metric threadform -- which are both 60-degree threads with a 1/8 Pitch flat at the Major Diameter and a 1/4 Pitch flat at the Minor Diameter -- can be calculated easily from the Thread Pitch:

Theoretical Single Depth = (5/8) x Pitch x Cos (30 degrees)

Working a numerical example for a Unified threadform of 1/20 inch Pitch:

Theoretical Single Depth = (5/8) x 1/20 inch x Cos (30 degrees)

= 0.625 x 0.05 inch x 0.866

= 0.0270 inch

Working a numerical example for an ISO Metric threadform of 1.25 millimeter Pitch:

Theoretical Single Depth = (5/8) x 1.25 millimeter x Cos (30 degrees)

= 0.625 x 1.25 millimeter x 0.866

= 0.677 millimeter

The Unified and ISO Metric threadforms are modified (truncated) V screwthreads with a 60-degree included angle between the flanks. In a 60-degree Sharp-V screwthread, the measured-along-the-flank length would be exactly equal to the Pitch. The flats at the Major and Minor Diameters on the Unified and ISO Metric threadforms result in a reduction of the measured-along-flank length to 5/8 that of a same-pitch Sharp-V screwthread. [This means that the Theoretical Along-Flank Infeed of a compound / top slide set at 30 degrees to feed along the flank can be calculated as (5/8) x Pitch, which for a Unified threadform is exactly equal to (5/8) / Number of Threads per Inch.]

The Single Depth of the screwthread is the measured-along-flank length of the screwthread times the Cosine of one-half of the V angle, i.e., 30 degrees.

I would suggest sitting down at your computer and making a spreadsheet for every possible TPI, or metric pitch, that you may be expected to cut, using the formula(s) above. Then when you are ready to cut some threads, you only need to know the nominal diameter and can look up the suggested diameters in Machinery's Handbook for the class of fit that you want.

++++

As an aside, if you cutting your threads properly, you don't need to file anything down, nor do you need to use Scotch-Brite on them, either.

 

[wquiles]

Will, as you know I advocate the use of the 29.5* method over straight infeed. However I would suggest hitting the peaks of the threads lightly with a file with plenty of WD-40 or whatever you use on the next to last pass. Then make your final pass of about .001-.002". Then spin the spindle up to 1000 or so RPM and hit it with a Red and then Gray Scotchbrite again soaked with WD-40.

OK, I will try the Scotchbrite pads - supermarket stuff?

29.5° is also what I use, as it's easy to set the compound to that angle. Then a fishtail gage is used to set the tool at 90° to the part being threaded.

For almost all lathe ops, carbide tooling is used. The exception is single point threading in aluminum. A carbide insert is never as sharp as a freshly ground HSS tool, and carbide tends to tear most alloys including 6061. It's easy to get the correct point angle, again using the fishtail gage.

I also use mag-back bases for threading. When using 29.5° compound, you make a pass & retract the tool, go back to the start position, advance the tool back to where it last was, zero out the dial, and dial in more depth. Much faster to do it than to write about it:thinking:

Yep, I used that tool to line-up the bit - alignment of the cutting tool is not the problem this time.

As to the actual process:
1) engage cariage at #1 on the thread dial
2) set cutter depth
3) turn on motor
4) stop motor on the cut groove
5) turn motor in reverse
6) once cutter is back past the piece, proceed to #2
7) once final depth is achieved, do a couple of pases
8) clean cuts with a metal brush

Note that the carriage is always engaged and I use the reverse on the motor to go back and forth.

 

[jhanko]

Your problem is caused by your reverse procedure. There is too much backlash in the change gears, lead screw, half nut, etc. to expect the cutter to be in the same position in reverse as it was in forward. Before you reverse the machine, back off the cutter to slightly clear the O.D. of the work. The cutter is rubbing the side of the thread in reverse, distorting the material. I learned this the hard way also..

Jeff

 

[gadget_lover]

Your problem is caused by your reverse procedure. There is too much backlash in the change gears, lead screw, half nut, etc. to expect the cutter to be in the same position in reverse as it was in forward. Before you reverse the machine, back off the cutter to slightly clear the O.D. of the work. The cutter is rubbing the side of the thread in reverse, distorting the material. I learned this the hard way also..

Jeff

This makes perfect sense, as in reverse you are not cutting, but are instead rubbing. That , combined with the backlash as Jhanko pointed out, could easily deform the threads.


Good call!

Daniel

 

[DUQ]

OK, I will try the Scotchbrite pads - supermarket stuff?

MM meant 3M hand pads. They come in different colors (grade)

http://www.3mestore.com/hand-pads.html

 

[ICUDoc]

Will I am often nervous cutting threads at 29.5 degrees as the carbide cutter I use seems to remove a tiny bit of the newly formed thread all the way from its base ( at the forming minor diameter) to its tip ( all the way to the major diameter) and have wondered whether it is actually making the MajDia smaller with the last few passes. it always seems to fit OK. Maybe this will prevent the "upsizing" you are seeing???
I am an amateur at this but I offer it FWIW...

 

[Mirage_Man]

MM meant 3M hand pads. They come in different colors (grade)

http://www.3mestore.com/hand-pads.html

Yup, Scotch-Brite Hand Pads.

Enco has them...

Red/Maroon Very Fine
Gray Ultra Fine

 

[PEU]

http://video.google.com/videoplay?docid=5557473837334990194

Yeterday I adapted one of my toolholders to accept my new threading tool, I dont believe how nice the thread resulted, best thread I did ever! It uses TPMC 16 inserts.


Pablo

 

[Mirage_Man]

http://video.google.com/videoplay?docid=5557473837334990194

Yeterday I adapted one of my toolholders to accept my new threading tool, I dont believe how nice the thread resulted, best thread I did ever! It uses TPMC 16 inserts.


Pablo

That's sweet! Wish I had a CNC machine.

 

[greenlight]

I love watching your demonstration videos.

 

[Anglepoise]

Will,
If you intend to make parts that have to fit other parts you have never seen or measured, I would take Fred's advise.

Also I don't see from your detailed post if you are reversing with the cutting tool still engaged with the workpiece. If you are doing this, the slack in the gears will completely screw up the thread (as posted by member JHanko). Always disengage the tool from the workpiece when winding back. Knurling and threading are a pain but your threads have to be perfect if the items is to fit an unseen piece as you can not trial fit.

 

[jhanko]

Looking at your pictures, I just noticed something else that will definitely cause problems although it has nothing to do with your increase in diameter problem. I noticed that as you are making each cut, you are cutting the feed in half. That is correct if your goal is reducing diameter as the amount you cut will result in a double that amount in diameter reduction. When cutting threads, The depth of feed is just that, not halved. The first cut you made should have been .008", but you did .004". When you were all done, you were at .019". A little more than half as deep as it should have been. To get a great thread on your mini-lathe, try this:

Follow that chart you have excactly. Don't cut the feeds in half. When you get to the last pass, keep repeating it until the cutter isn't removing anything anymore. This will vary depending on how loose the lathe is.
Example for 20 TPI: .008", .007", .006", .005", .004", .003", .000", .000", .000", .000". You will notice that even though you make 4 or more passes with the feed set at .033", it is still removing material, less and less each time. I usually make 4 passes at my final depth. When finished, all you'll need to do is wipe off with a terry cloth towel and admire the mirror like finish. As PhotonFanatic stated, when finished cutting the threads, you shouldn't have to touch them with anything (file, sandpaper, scotchbrite pad).

Most importantly, never let the cutter touch the work while reversing.

Jeff

 

[wquiles]

Too hard to thank each one individually - thanks so very much for the many replies/suggestions. :thumbsup:

I will be trying again tomorrow morning (garage is now at 101F here in Texas!!!) and will update post #2 with my results as it is obvious I am doing not one, but several things wrong :shakehead

Now, I do seem to have some different goals for thread depth: Fred's equations state a thread depth of 0.0270", but JHanko/thread table states a depth of 0.0330". I think I will try 0.027" first and see how it works out

Again, you guys are awesome - thanks much for the help :twothumbs

Will

 

[PhotonFanatic]

Will,

You will need to re-read what I posted--if you are like me, you might need to re-read it about five times.

The .027" is the depth of thread for a straight infeed of the cutting bit into the material, i.e., at a right angle to the axis of the workpiece.

But, if you do switch to cutting using the flank infeed method, i.e., set the compound slide to 30-degrees, then the distance of infeed along the path that the tool bit will follow when using the compound slide is, in fact, .0312".

This graphic should make it a bit easier to visualize:

Hope that helps.

BTW, different depths of cut are to be expected from different sources--one thing we aren't delving into here is the class of fit, or the % of thread engagement. But I believe that if you use that basic formula for depth of cut using the 30-degree setting, i.e., 5/8 / TPI, you will have at least 75% thread engagement and that will be sufficient for most flashlight pieces.