need help on double start threads
In all my years I have never had to make a double start thread until now .
Now my old planer stripped the cross nut and work is stacking up .
I need to get this fixed fast !
The nut is a brass insert that has 1 1/8- 16 OD threads- thats the easy part .
Cleaned up the screw and found out it is a 3/4" -2 1/2' double thread .
( measured 2", counted 10 thds , so 10/2= 5/2=2.5tpi )
my lathe will cut 2.5, but how do I get them to start 180 deg apart?
Is this something as simple as engaging opposite marks on the chasing dial ?
I have a choice of my newer china lathe , or an older "Advance" American iron.
The Advance has a lever to disengage a 4 way coupling that drives the lead screw.
Is this the key to re sychronize opposing thread starts ?
Cut a few passes , stop , disengage dog clutch , roll by hand to get opposite dogs ?
This recent thread has some good info. Thread dial method is discussed, as is using the compound, with it set to 90*, then offsetting the tool with that and leaving the halfnut engaged or using the same position on the thread dial each time. Sorry, but I cannot provide any personal experience.
I have cut double and quadruple lead threads, with the compound set at 90 degree method. it works well. cut a sample part first to check your lead
If your threading dial is like the one on the Monarch EE, this is easy. Normally, when chasing a single 2.5 tpi thread, you'd engage the half nuts when the thread dial was on 1 or 3. Is your machine the same? Successive cuts would have to start with the thread dial on 1 or 3, right?
Go ahead and take a scratch cut that way.
Now, to add the second thread, engage the half nuts on either 2 or 4.
Done. It's that simple. Doesn't always work out this nice - depends on the pitch.
10/2 = 5/2 = 2.5 tpi??????? Something wrong here Bob.
Originally Posted by FlatBeltBob
I'd have to agree. If you count ten threads in one inch and you're certain there are two starts, the lead is .200"= 5 threads per inch.
Originally Posted by luthor
Whats wrong with my math?
if I count 10 threads in 2" then there are 5 in 1 inch , but if there is a double thread , then there are only 2 1/2 threads per inch , with two threads next to each other .
Maybe I should have said there are 10 thread crests in 2 " , but they are 2 seperate threads .
If I divide1" by 2.5 I get a pitch of .400"
So if I engage the half nut and leave it engaged , I will take a few passes , reversing the spindle to get to the start point each time .
Then with my compound set at 90 degrees , I advance the tool by .400 and repeat .
( I have an indicator setup to measure compound movement ) .
If I alternate between the 2 threads , and finish them at about the same rate , I can sneak up on the finish size , till the lead screw fits properly .
Is this correct ?
I forgot to mention that the thread form is not an acme . it is more of a square thread form .
I plan to grind the tool and use the external thread as a gage .
Measuring the OD of the thread , and depth of thread should get me close to the root dia.
Then I plan to bore a short ID at the screw OD dimension to use as a visual gage for depth of the internal thread of the nut .
ok, a bit of visualization will be required here, and I have done this before, once, and would do it slightly differently again, but I digress. 1. set the compound as you would for a single lead thread, be it 29, 14, etc. Calculate a triangle of how far to retract the compound and compensate with the cross slide to shift the pitch 1 lead. The compound travel, the hypotenuse, the lead the sine of the angle of the compound. Cut one lead and measure it as you wish, wires,, mics, whatever. When you get it where you think you want it, spray the whole thing in layout dye, shift the compound 1 lead and cut the next. The dye shows you which lead you are cutting. On a rectangular thread make sure your tool is a couple to 5 thousandths wider than you think you need it, especially on a worn thread, as the web will vary between leads in worn areas. Using the compound method allows you to shift back as well. Using opposing numbers on the chase dial might work for standard threads, but this method will even work when the halfnuts have to remain engaged ( most metric). Another method is an index plate on the chuck, and basically dog the part against a different pin in an index plate to shift lead position.
Double start thread
An easy way is to use a four jaw chuck and just index the start of the threads off two opposing jaws. For a triple start thread use a three jaw chuck. Sometimes simplicity is the key.
Done it the "easy way" before myself, but by putting the work between centers and using the slots in the dog driver plate as my index. Anything can be done with a little thought, but boy, hasn't CNC ruined us? Just enter a value for the number of leads and go with it...
I helped a Bud around the corner doit,I had him use a felt pen to get his couage up,shaft was aboot 4" in dia.
On his lathe,it was simply oppisite numbers on the thread dial.
Assuming all else said is correct, and lead is truly .400", then only .200 (actual movement) on compound. Moving it .400" will cut in the same groove.
Cutting tool will have to have the clearances needed. I.E., assuming a thread "pitch diameter" of about .650", helix angle will be in the neighborhood of 11 degrees.
The killer on any "nut" of this type will be the teeny diameter boring bar needed as compared to length of nut.
Last edited by johnoder; 02-16-2010 at 05:31 AM.
Reason: add info
I did a sample thread with turning the compound 90 degrees, or parralel to the Z axis. Its a simple setup, and universal for what ever number of multi-start threads you will cut. Basically, since you have a two start thread, you advance the compound one pitch to cut the next thread. I cut the first thread near depth, then cut the other thread near depth. Then I did the finish passes. I also put a dial indicator on the compound to get an accurate measurment of its movement.
Yes. See my post #4.
Originally Posted by FlatBeltBob
Yes , you are correct . I played around a bit today with the screw chucked in the lathe .
Originally Posted by johnoder
Cut a few 'phantom' threads and sure enough , if I reset the tool .200" using the compound , I pick up the 2nd thread . Forget about alternating the chasing dial marks because this lathe just won't cooperate that way .
So the plan is to engage the half nut and keep it closed till both threads are done .
Since this is NOT an acme thread , but rather a square profile it is possible to get the correct depth , and still have a tight fit . Being able to side shift the tool using the compound will allow me to sneak up on the fit by shaving the sides of the thread .
For sure , this will require a small bar , but I have done this before , just not on a double start . This is a good job for after hours . Doors locked , phone off the hook .
I want to thank every one for all the info . Will let you know how it works out.
The vertical screw will probably get inspected next .
This 100 year old machine is still earning its keep as a daily runner .
Mostly grinding wood chipper knives . I just need to kepp it runnin another 20 or so ...
In your 1st post you quoted 10/2 = 5/2 = 2.5tpi, since when did 10/2 = 5/2?
Originally Posted by FlatBeltBob
See posts 7 & 8 .
I should not have expressed it as an equation .
In post 8, I said I should off set by .400" to cut the 2nd thread , but as John Oder explained , that would merely jump me to the next crest of the same thread .
The correct offset is .200.
This article helped me understand how to do it. I cut them fairly often now as a lot of the bigger gate & globe valves I repair at work have double start acme threaded nuts on the valve stem. I use the thread dial when the pitch allows, and when I can't do it that way I'll cut one thread complete, throw my spindle & gearbox into neutral and index the next thread 180deg out. It's easier to learn cutting an external thread and I practiced the first time buy replicating the valve stem I was making the nut for. Made for a nice go-gage since the valve was too heavy to lift for checking fitment without disassembling.