Measuring TIR

T

tigtorch

Active User
Registered
I recently decided to try to make my own spindle mounted (1.5X 8 tpi) ER40 collet chuck for my Rockwell 10 lathe. I bought a kit from Jan at Tallgrasstools.com and managed to make it, his drawings and instruction were very well done. One of our members also helped me a bit via PM's (Stevecmo). It was a very rewarding experience because until this project I had never successfully even threaded anything with a lathe. Since I do not have metric threading capability, I also made my own ER40 nut.

Once done I wanted to measure the TIR and thus my question: Until I successfully finished this I didn't want to buy an entire collet set so the only one I own is a 13/32 ER40 collet I bought used from Ebay. My first attempt at measuring TIR was to turn down about 1.5" of 1/2" steel rod in my 3-jaw chuck to about 0.410 to be "clampable" by the chuck. I then chucked this piece of steel in the collet chuck and placed my indicator on this rod, near the chuck. This resulted in about 0.002 TIR. I then took a piece of aluminum rod, turned it smooth, chucked it in my 4-jaw and got it dialed in so that the runout was un-measurable. I then turned this down to 0.410. I then chucked this into my collet chuck and the TIR was nearly un-measurable, maybe 0.0005 or less.

Why was the first effort so much worse than the second? I realize my 3-jaw chuck is probably not very good TIR wise but once turned, the piece of steel rod shouldn't care, should it? My guess is that the steel flexed in the lathe (the finish wasn't that great, I probably should have attended to the HSS bit a "bit" better, plus I am not even sure of the type of steel it is). The aluminum, on the other hand, had an almost mirror finish. I would really like to learn from this if anyone can shed any light on it!
 
I agree with your assumption - the cuts on the steel were too deep making the part flex and pulling it out of round which produced your TIR indifference.

how light where the cuts on the steel?

as a general rule of thumb, when making any measuring devices/bars/etc always take very light cuts to produce the best finish you can. this is one of those cuts where it really matters more than others.
 
Could be a few things going on.

How long was the piece of steel and how much of it did you surface? When you put the piece in your collet, did you DI off an area that you surfaced?

Anyhow, a 1/2" steel rod protruding 1.5" from the jaws will bend a little. There's a trade-off. If you cut deep enough to get a good finish, it applies more pressure and bends the workpiece (thus, a center is usually used). If you take very light cuts, you'll sacrifice the finish quality and you won't be able to DI off the surface worth a darn.

AL cuts like butter and makes you feel like a hero just like when you smash a beer can after slamming-down the contents.

Also, a typical AL round-stock from the store has a nearly perfect finish -that's never the case with steel (unless you purchase TGP condition steel). If you were DI'ing off the factory finish, that could explain a lot.

Ray


tigtorch said:
I recently decided to try to make my own spindle mounted (1.5X 8 tpi) ER40 collet chuck for my Rockwell 10 lathe. I bought a kit from Jan at Tallgrasstools.com and managed to make it, his drawings and instruction were very well done. One of our members also helped me a bit via PM's (Stevecmo). It was a very rewarding experience because until this project I had never successfully even threaded anything with a lathe. Since I do not have metric threading capability, I also made my own ER40 nut.

Once done I wanted to measure the TIR and thus my question: Until I successfully finished this I didn't want to buy an entire collet set so the only one I own is a 13/32 ER40 collet I bought used from Ebay. My first attempt at measuring TIR was to turn down about 1.5" of 1/2" steel rod in my 3-jaw chuck to about 0.410 to be "clampable" by the chuck. I then chucked this piece of steel in the collet chuck and placed my indicator on this rod, near the chuck. This resulted in about 0.002 TIR. I then took a piece of aluminum rod, turned it smooth, chucked it in my 4-jaw and got it dialed in so that the runout was un-measurable. I then turned this down to 0.410. I then chucked this into my collet chuck and the TIR was nearly un-measurable, maybe 0.0005 or less.

Why was the first effort so much worse than the second? I realize my 3-jaw chuck is probably not very good TIR wise but once turned, the piece of steel rod shouldn't care, should it? My guess is that the steel flexed in the lathe (the finish wasn't that great, I probably should have attended to the HSS bit a "bit" better, plus I am not even sure of the type of steel it is). The aluminum, on the other hand, had an almost mirror finish. I would really like to learn from this if anyone can shed any light on it!
Click to expand...
 
Your assessment was what I was thinking too. The steel cuts harder and you have more push-away. Many times when am turning steel, I leave a little groove or under-cut at the end of the cut, so when I am at the end of the cut on the final pass, the tool is resting in the groove. Don't touch the cross-slide or compound handles or back them off. Then I shut of the spindle and slowly crank the saddle back. There is usually a slight scratch because the steel pushed away a little. I then turn on the spindle again and make a final "free" cut so the push away is eliminated. Rich
 
Thank you guys for your quick responses. I didn't take less than a .010 cut on the steel, even though I was only cutting about 1.5" it was sticking out at least an inch more than than from the chuck and as I said the bit I was using I guess wasn't sharpened as good as it should be. I think this answers my question. Can I be reasonably confident my collet chuck is decent?
 
tigtorch said:
Thank you guys for your quick responses. I didn't take less than a .010 cut on the steel, even though I was only cutting about 1.5" it was sticking out at least an inch more than than from the chuck and as I said the bit I was using I guess wasn't sharpened as good as it should be. I think this answers my question. Can I be reasonably confident my collet chuck is decent?
Click to expand...

Personally - for a measuring bar I wouldn't take cuts heavier than .005 - with the finishing cuts being around 0.002 , not so much for the depth of cut as much as it's to reduce pushing forces on the bar being cut (flexing issues). Also as mentioned - I would center drill and use a dead center (not live) and lube to add stability to these cuts.

I think your chuck is solid from what you've written.
 
... Welllll.... It certainly means you're headed in the right direction but, you can't assume the only collet you have is dead-on. Matter of fact, in the absence of some known proof rods and some creative DI measurements, it will be challenging to find-out exactly where you stand.

I have no experience with ER collets but, assuming they're like any other workholding tool, you have to gain experience with the entire range of collets before you know where you stand. -That might involve spot checking different ones with known proof rods of different diameters.

The fact that you're running within a half-thou is a really good sign that you did pretty good. -Congrats.

Ray



tigtorch said:
Thank you guys for your quick responses. I didn't take less than a .010 cut on the steel, even though I was only cutting about 1.5" it was sticking out at least an inch more than than from the chuck and as I said the bit I was using I guess wasn't sharpened as good as it should be. I think this answers my question. Can I be reasonably confident my collet chuck is decent?
Click to expand...
 
Hello tigtorch,
I'm not clear on the procedure you used.
When you completed the turning in the 3-jaw, the TIR on the turned portion should have been at the level of the spindle bearings accuracy before it was removed from the 3-jaw.

When you removed the part from the 3-jaw and chucked the turned portion in the collet, did you:
1)indicate against the turned portion, or
2)indicate against the unturned portion that had been in the 3-jaw?

If 1) then you should get TIR consistent with the collet & spindle accuracy (and you may be experiencing the issues that you and others have mentioned)
If 2) then you are probably measuring some of the runout of the 3-jaw

Terry S.
 
Twstoerzinger, I am a novice but I am not stupid! I don't mean to be rude but if I did option 2 and didn't mention it or realize it's wrongness, I don't think I could have made the chuck.
 
You must log in or register to reply here.
Back
Top