Mazak QT 20 chatter problem

[mavmike]

Ive been running this machine for 12yrs with no problem then all of a sudden. About a week ago i started noticing chatter on all my parts on the OD. I work in a job shop so im cutting all different sizes. It doesn't matter if its 3" or .5 OD its gonna chatter. You may get lucky and one or two out of ten won't. I no its not feed and speed because ive worked with the same material for a long time. I mostly machine A2 tool steel and 4140 ph. Im thinking maybe spindle bearings. What do you all think? Thanks so much Mike

[mavmike]

Anybody?

[mavmike]

No ideas???????

[girishdk]

Hi,

Did you check the backlash for X, Z axes? Does this machine have linear scale or only motor encoders?

[CNCRim]

I don't think I can help much...

OD only, does it chatter on ID? If chatter is on OD only not ID I suspect X axis has backlash.

[Leelund]

Check and see if the machine uses gibs for X axis and Z axis as well. Maybe they need adjusted. Hows tool height? Maybe an unexpected bump? If it's not tooling issue it's machine and something has moved or has loosened. If you think it's the spindle bearings check end play and runout with an indicator.

[service_bb]

Could be in the turret. Ran into simliar problem a while back. Chips were behind the turret, so when it clamped, it didn't have suffcient clamp pressure qnd caused chatter.

[mavmike]

We had the turrett encoder replaced about 3weeks ago. The turrett would spin like a helicopter and lose position. Do you think it could be related to my problem? Ive done some ID work and got a good finish. What is the best way to check backlash on the x axis. Go figure my company laid off our maintance man so ive got to work on this thing. O boy this is gonna be fun. Thanks for the help so far.

[jimiscnc]

A journeyman cnc tech has the experience and knowledge on how to approach this problem. And I ain't a service tech.

However, there's a few clues to look for assignable cause variation:

I.D. has no chatter - cutting forces in X are pushing down on the turret. (X minus direction)

O.D. the cutting forces are pushing the turret up. (X plus direction) this could bring out thrust bearing issues on the ballscrew or a ball nut worn in the region everything is when you're doing your O.D. turn?

The turret having been recently taken apart could be a factor?

Finally, on most CNC lathes, they happen to be much stiffer when the turning force is going into the bed. This would be clockwise spindle rotation as the operator views the work area.

CCW turning tries to lift the entire turret up and that's opposed by the gib straps, which is usually less surface area and thus less strong that CW turning.

I saw this scenario once with my own eyes. Chatter when CCW, smooth finish when CW.

FWIW

-90% jimmy

[mavmike]

In regards to the direction of cut my rough turn tool is upside down and finisher is facing up. There is chatter on both tool going either direction. The thrustbearing and the thrust nut is something ill check. Is this a pain to check and also how do i check backlash. Thanks

[mavmike]

Also the turett was not removed all we did was remove rear cover to install new encoder.

[jimiscnc]

backlash/reversal error/ dead zone/ lost motion. all synonyms for the term backlash to me, anyway.

Here's a suggestion. (Like I said, I am not a tech, so I wish they would chime in.)

put indicator on turret. have stock in the spindle. keep the spindle fixed - by M19 if possible.

using the pulse generator, go x minus until the indicator reads zero against the stock in the spindle. turn the res on the pulse generator (handwheel) to .0001" per click. handwheel in X plus and compare the number of pulse cliks it took to get the indicator needle to move. this is your backlash.

if it's ballcrew/ball nut, you may get differing readings depending on what part of the ballscrew you're looking at. In other words, the most wear is in the region of the ballscrew stroke that is used the most for machining. Extreme ends of travel would show less wear.

With the pulse generator, to the best of my knowledge, it doesn't use reversal error comp and pitch error comp. These comp factors are put into machine motion when movement is by program, however.

The second experiment would be to remove the guarding enough to get at the thrust bearings for the ballscrew. See if you can see or use an indicator to see if you can spot any reversal error from the ballscrew to the thrust bearing. I'm not sure if there's a thrust bearing on both ends. I think the thrust bearing is at the end up of the slide with the servo on it at the highest part of the slide.

Sorry for the vague generalities. I'm just a damn country Doctor, Jim!

spindle bearings, worn ball nuts, worn ballscrew, gib/way wear and adjustment, spindle drive belt tension, chuck gripping force loss at rpm (chuck should be lubed DAILY) and a couple of other things could be a factor - clipping a lube line when replacing the turret encoder could even be a problem. boring jaws straight without back taper could also do it, believe it or not! Even with a brand new machine I would usually bore a back taper of about .001" on the dia. per inch of depth. So my jaw boring program would start at 2.000 and at 1" deep, the programmed dia would be 2.001". This is because all chucks have a "diaphramming effect due to the centrifical force on the jaws at RPM. If you bored 2.000 for example, if you could measure your bore at 2500 RPM without holding a part, the real diameter at speed may be 2.004" at the front and 2.000 at the rear shoulder of the bored jaws. This gives you POINT CONTACT for gripping instead of the cylindrical contact you really need.

I saw once in my life a turcite lined vee way in X axis on a VMC completely loose it's accuracy. lasered it again, calibrated it to the best of our abilities and still junk accuracy. The ultimate culprit was BRAND X WAY OIL! It had to be Mobil or Shell Tellus brand for the machine to get it's accuracy back. lube flush and correct waylube brought the one year old machine back to good as new!

Another thing for spindle bearings is to put an indicator on the chuck and lift the spindle with a piece of wood. If any indicator movement beyond .0003-.0006" or so is cause for concern.

Machine tool spindle bearing replacement is an art. You need to surface grind bearing shims and if you even touch the bearings with your fingers it's instant scrap.

There should be some helpful diagrams for dealing with this stuff in your machine maintenance manual. The pictures and views can prove helpful in getting at stuff without tearing your machine completely apart.

This is boilerplate for a service tech. I only know what I know by looking over their shoulders all these years. Big diff between doing it yourself and merely watching somebody else do it.

jim