Is my indicator damaged?

[Lucaselef]

I have 2 interapid indicators one .0001" and one .0005". I use them all the time and really trust them. I recently upgraded to number of noga mag bases, the steel arm and also the articulating type. My previous Chinese articulating type had the the adjustment at the end of the arm and I would constantly get flutter when measuring.

Well I set up to dial in a chuck backplate with the beefy steel arm noga base, and the interapid .0001" indicator. The backplate has a very nice surface finish. When I started trying to dial it in as close to .0001" runout as I could get I noticed the indicator was drifting clockwise after the part stopped moving. I was able to replicate the drifting with both indicators, all 3 noga bases, the backplate as well as a ground 1/2" gauge pin.

Here is a video of the needle "drifting". You can see as I stop rotating the chuck the needle drifts a few tenths clockwise.

Is this somewhat normal at this measurement level, or are my indicators damaged? I don't think I ever previously noticed this phenomenon as my previous base wasn't stable enough and I'd always have a little needle bounce going on. That, or the surface roughness of the part would induce needle bounce on its own.

Any ideas?

 

[Chris Lesiak]

I think it is just the hydrodynamic lubrication changing in the plain bearing headstock.

 

[Lucaselef]

I think it is just the hydrodynamic lubrication changing in the plain bearing headstock.

I was afraid it might be that. I tried moving it very slowly to keep from making the spindle ride up on the oil and it seems to help a little, but still frustrating.

Anyone have any tips for dialing in stock very precisely with a plain bearing headstock, or am I just stuck with ~.0005" alignment?

I tried running the spindle with the indicator on the gage pin at very slow speed, but have to use back gears to get the slow speed and that introduces quite a bit of noise.

 

[eKretz]

I was just about to ask if that was a plain bearing headstock. Chris probably nailed it. Check your clearance and adjust if necessary. That's about all you can do. I would just compare your settled readings to your running ones and see what you get. In other words, if the settled readings say you are zeroed, run the headstock slowly and see if they stay there but just shift a little (I.E. instead of zero at every jaw, you see +.0003" at every jaw). If so, you can pretty safely ignore that. If not, I'd maybe try to devise a way to indicate with the spindle moving. Maybe a hand crank you could use on the end of the spindle? Don't suppose you're running 3 phase, eh? A VFD would be perfect.

 

[Lucaselef]

I was just about to ask if that was a plain bearing headstock. Chris probably nailed it. Check your clearance and adjust if necessary. That's about all you can do. I would just compare your settled readings to your running ones and see what you get. In other words, if the settled readings say you are zeroed, run the headstock slowly and see if they stay there but just shift a little (I.E. instead of zero at every jaw, you see +.0003" at every jaw). If so, you can pretty safely ignore that. If not, I'd maybe try to devise a way to indicate with the spindle moving. Maybe a hand crank you could use on the end of the spindle? Don't suppose you're running 3 phase, eh? A VFD would be perfect.

The bearing clearances we're set not long ago when I rebuilt the lathe. They're set almost as good as they can get on the lower end of the recommendation by southbend (it's a heavy 10 lathe). I can't remember exactly what the spec calls for, but I think it was like .0007 or .0008" total, or .00035-.0004 per side. Which would explain the couple of tenths sagging as it pushes the oil out of the way while stopped.

Maybe I can figure out a way to mount my 5c hand collet closer just to help spin the chuck.

Unfortunately I'm running a 1hp single phase motor right now. I'd like to put a 3 phase inverter duty motor and vfd on it, but since I have a threaded spindle I wouldn't be able to stop quickly or reverse quickly so it'd be a lot of money just to vary the speed. I'll most likely do it at some point in the future, but it's not in the cards right now.

 

[DDoug]

Switch to a .001 indicator, and get on with work.

 

[michiganbuck]

Switch to a .001 indicator, and get on with work.

Agree to the .001indicator.. If you make/shim a plane-bearing spindle too tight it may begin to rub when you run at the highest RPM.

Some grinding shops would send in a start man to fire up grinders to warm them up.
It is the same with a lathe..when fully warmed up they are tighter ..but too tight is a big problem.

 

[dana gear]

Funny tidbit, My CNC lathe decided it needed a vacation, while it was resting, I sent some work off site to a small one-man machine shop, He had an older Japanese CNC lathe with a bar feeder. He would leave his CNC lathe running all night (auto chuck spinning) when he was gone, as he told me that he would have to wait a couple hours in the morning to allow it to warm up, so he just left it going, he said it would hold tolerance better when warm. Hay whatever works for ya, He did do excellent work.

 

[jccaclimber]

Funny tidbit, My CNC lathe decided it needed a vacation, while it was resting, I sent some work off site to a small one-man machine shop, He had an older Japanese CNC lathe with a bar feeder. He would leave his CNC lathe running all night (auto chuck spinning) when he was gone, as he told me that he would have to wait a couple hours in the morning to allow it to warm up, so he just left it going, he said it would hold tolerance better when warm. Hay whatever works for ya, He did do excellent work.

Place I used to work at had a field of Mazak Dual Turn 20's. Robot feeder, but it pulled from a tray so still a finite run time before attention was needed. By the end of break some of the machines would end up idle for 10, sometimes 20 minutes depending on when they were loaded (multiple machines per operator). It was enough idle time that everything would be out of whack by a few tenths for the first couple parts (3-5 minute takt time). They had a warm up program they could run. It was the exact same as the production program, it just ran without the part in the chuck, so a few alarms had to be disabled. It helped, but even at the same speeds, without actual load on the spindle we could see things move.