Follow along with the video below to see how to install our site as a web app on your home screen.
Note: This feature may not be available in some browsers.
Largest Manufacturing Technology Community on the Web
Stay Connected:
Two Collar Test Question
- Thread starter Axl
- Start date
- Replies 84
- Views 8,981
Limy Sami
Diamond
- Joined
- Jan 7, 2007
- Location
- Norfolk, UK
Yes - less wear on the tool.
atomarc
Diamond
- Joined
- Mar 16, 2009
- Location
- Eureka, CA
Or...I want to see the difference between the headstock end and the tailstock end, and the mid point between theses will be self evident...or not important.
Stuart
Stuart
some make it a "three collar" test. One halfway between, to check for non-linear taper.
The less wear on the tool issue is mostly to make sure the "same tool" works on all the "collars".... which it won't be if it wears down a bit between them. I don't know how real a concern that is, but at least you know it can't wear if there is nothing there for it to have to cut. Gets rid of an objection.
The less wear on the tool issue is mostly to make sure the "same tool" works on all the "collars".... which it won't be if it wears down a bit between them. I don't know how real a concern that is, but at least you know it can't wear if there is nothing there for it to have to cut. Gets rid of an objection.
imo unnecessary if doing a test at the chuck. the bar loses rigidity or you lose time cutting the relieve multiple times. how much does an insert wear over 150 - 300 mm? if adjusting the tail it might save some time. or maybe not, engaging, disengaging, cranking. why not wait for a few seconds? if the bar is too long you have push off anyway. also weird things might happen when the tool enters the work.
EPAIII
Diamond
- Joined
- Nov 23, 2003
- Location
- Beaumont, TX, USA
Perhaps I have looked at it a little differently. The important thing in this test is that the two diameters be exactly equal. Now if you just turn a large cylinder, trying to have the same diameter from end to end, and your lathe is not known to be set up exactly right, then you are likely to have a slight taper.
By removing some of that questionable area between the two ends, you make it easier to know that the two, narrow areas which are left are the place where you want to do your measurements. You would turn the complete OD. Then reduce the OD between the ends. Then finish the two ends using your best micrometer so their diameters match as close as you can get them. When you use the test bar, you know those two ends are the same diameter and you are not on a tapered area between them because that area has been cut off.
By removing some of that questionable area between the two ends, you make it easier to know that the two, narrow areas which are left are the place where you want to do your measurements. You would turn the complete OD. Then reduce the OD between the ends. Then finish the two ends using your best micrometer so their diameters match as close as you can get them. When you use the test bar, you know those two ends are the same diameter and you are not on a tapered area between them because that area has been cut off.
Tyrone Shoelaces
Diamond
- Joined
- Apr 19, 2006
- Location
- Manchester, England
I've known operators doing this procedure on a DRO equipped lathe cut the band at the open end, zero the DRO, come out and move to the other end. They plunge in gently back to zero, then cut another band.
Obviously they are only taking the lightest of cuts and the " plunge in " is done very carefully.
Regards Tyrone.
Obviously they are only taking the lightest of cuts and the " plunge in " is done very carefully.
Regards Tyrone.
steve-l
Titanium
- Joined
- Mar 2, 2012
- Location
- Geilenkirchen, Germany
This subject has come up many times in previous threads. I see no value in the test, as it delivers that there is or is not a taper. It does not deliver the cause. If a taper is being cut, the cause can be a whole host of different reasons. None of which can be identified with just this test. If you don't know the cause, how can you correct the fault? The answer is the use of a test bar that mates to the lathe spindle taper and a DI. The test is both faster to execute and definitive. Test bars are inexpensive and if you can afford the lathe, you can afford the $100 for a test bar.
Peter from Holland
Diamond
- Joined
- Jan 15, 2005
- Location
- The Netherlands
But now comes the part where you have to adjust the head
I have done it with 2 indicators One on each collar both set to 0
Then you start adjusting Both indicators move
Now the head is correct if the DIFFERENCE between those indicators is the difference you measured on the RADIUS
Peter
I have done it with 2 indicators One on each collar both set to 0
Then you start adjusting Both indicators move
Now the head is correct if the DIFFERENCE between those indicators is the difference you measured on the RADIUS
Peter
cyanidekid
Titanium
- Joined
- Jun 4, 2016
- Location
- Brooklyn NYC
HEAT,(yes Demon73) and of corse to save time, more than tool wear(although good to reduce that as a parameter). as the part heats up, as it will from the cut, even if its a light cut (less so with coolant, obviously), and also from the friction of rotating on a center, the test bar expands in diameter, so a bar cut end to end, all other things being ideal, one would expect to have a taper upon cooling.
obviously, the actual amount of that taper depends on a number of factors, ( material of the test bar, diameter, temperature change during the cut, etc.), but if the heat from your fingers throws a micrometer off, well the heat of the machining operation is going to have an effect, no?
obviously, the actual amount of that taper depends on a number of factors, ( material of the test bar, diameter, temperature change during the cut, etc.), but if the heat from your fingers throws a micrometer off, well the heat of the machining operation is going to have an effect, no?
Peter from Holland
Diamond
- Joined
- Jan 15, 2005
- Location
- The Netherlands
You do not save time as you leave the feed engaged till both collars are done
Also no center involved on a 2 collar test
Peter
Limy Sami
Diamond
- Joined
- Jan 7, 2007
- Location
- Norfolk, UK
Something else not mentioned, ..yes you can have a $$$$$$$$$$$$ test bar, but you are measuring with that, whereas a turned test bar gives the result for actually cutting .........which are two completely different animals.
steve-l
Titanium
- Joined
- Mar 2, 2012
- Location
- Geilenkirchen, Germany
Of course you are correct, but having confidence that your machine is straight points you in the right direction for correction and test bars are NOT expensive.
jim rozen
Diamond
- Joined
- Feb 26, 2004
- Location
- peekskill, NY
" I see no value in the test, as it delivers that there is or is not a taper. It does not deliver the cause."
Then why did the manufacturer state that this test is the ultimate final arbiter of proper lathe setup?
Yep, read that how to run a lathe book.
Elaboration:
This test is used for one thing, and one thing only. It permits an owner to wring as much accuracy from a
worn lathe, as they possibly can.
1) lathe is used, bed wears.
2) this causes the machine to cut taper.
3) new operator performs two-collar test. Collars differ by 0.003 inch over 8 inch length.
4) adjustments are to either use the bed twisting screws on the newer SB machines or to put shims
under the tailstock legs on the older ones.
5) after an hour of testing and shimming, the machine now cuts collars within 0.0005 inch over the
8 inch bar length.
6) please note the tailstock was removed from the bed and set aside before this all started. NO tailstock.
7) the lathe bed is still worn. Outside of the 8 inch envelope it still turns taper.
8) you can't make an old lathe brand-new again. But you can adjust it to do what you want.
Two collar test. A piece of aluminum, a dead sharp HSS tool, and a micrometer. Test bar does not
help for this goal. Hi-zoot superduper level from starrett, does not help. Lasers do not help.
The machine either turns and bores straight, or it doesn't. That's the test.
Then why did the manufacturer state that this test is the ultimate final arbiter of proper lathe setup?
Yep, read that how to run a lathe book.
Elaboration:
This test is used for one thing, and one thing only. It permits an owner to wring as much accuracy from a
worn lathe, as they possibly can.
1) lathe is used, bed wears.
2) this causes the machine to cut taper.
3) new operator performs two-collar test. Collars differ by 0.003 inch over 8 inch length.
4) adjustments are to either use the bed twisting screws on the newer SB machines or to put shims
under the tailstock legs on the older ones.
5) after an hour of testing and shimming, the machine now cuts collars within 0.0005 inch over the
8 inch bar length.
6) please note the tailstock was removed from the bed and set aside before this all started. NO tailstock.
7) the lathe bed is still worn. Outside of the 8 inch envelope it still turns taper.
8) you can't make an old lathe brand-new again. But you can adjust it to do what you want.
Two collar test. A piece of aluminum, a dead sharp HSS tool, and a micrometer. Test bar does not
help for this goal. Hi-zoot superduper level from starrett, does not help. Lasers do not help.
The machine either turns and bores straight, or it doesn't. That's the test.
I firmly believe in the test as being an accurate means of checking, it has come straight from a South Bend set up manual. However I am definitely not counting out the test bar, as it is just that. Precision level is an excellent start point and the collar test should prove as a second setup method.
You are 100% correct, it definitely does not determine the cause. Thanks for your input!
You are 100% correct, it definitely does not determine the cause. Thanks for your input!
Monarchist
Diamond
- Joined
- Oct 2, 2012
- Location
- Sol, Terra
I "firmly believe" that SOMEBODY must be getting blow-jobs or 'air miles' if we can but resurrect this argument for the 100th time.
All it has proven is that human opinion is far more rigid than cast-iron lathe beds.
Are we getting close yet? 'Nuther five or ten years?
It is all in PM's database.
All it has proven is that human opinion is far more rigid than cast-iron lathe beds.
Are we getting close yet? 'Nuther five or ten years?
It is all in PM's database.
Tyrone Shoelaces
Diamond
- Joined
- Apr 19, 2006
- Location
- Manchester, England
