Poor Surface Finish on Lathe

[metronorth]

We are using a Colchester 220M lathe to turn an internal spherical radius. This is a part that we have been making since 2006 and have made many thousand times. Recently we have been having surface issues when turning the internal spherical surface. The part is 6061 T6511 aluminum, we have historically used 250 m/min surface speed with 2500 rpm as an rpm limit. We limit the rpm so that excess rumble is not generated by any imbalances in the machine spindle. We finish at .03 mm DOC and a feed rate of .03 mm / rev. We use a Sandvik VCGX H10 insert with a 0.8 nose radius. The surface finish required for this part is 0.4 um Ra. I am attaching a picture of the badly stepped surface finish that we are currently suffering with. We are slowing down our feed rate even more (30%) and are able to achieve the required surface finish but we obviously don't want to slow cycles. We have had the machine inspected for any mechanical deficiency (none found) and have also had a ball bar test done on it and have had the servo's optimized but there was no difference. Any ideas appreciated.

Glenn at Metro North Machine

 

[ifixcnc]

Finish Problem

What did "machine inspected for any mechanical deficiency" include? Spindle bearings, Axis backlash?

 

[HuFlungDung]

How does the machine perform when doing flat facing or turning? Any change to the surface finish then? Is the spindle getting noisy?

I had a strange experience several years ago, when one of my guys was doing some machining when suddenly the work 'jumped' position by quite a bit. Took the cover off the headstock and found that the spindle bearing preload nut had backed off several turns of thread and the bearing decided that was the day it was going to pop loose. Up to that moment, we hadn't seen any issue. But mysterious stuff happens once in a while.

 

[TDegenhart]

When did this start to happen, all at once or gradually over time? From the looks, it appears to be uniform in both in x and z.

To me it looks like a problem in the controller, that for some reason the pulses to the servos have coarsened or the positional feedback has changed. Do you have another machine with the same controller that you could put a scope on and on the machine giving problems?

Tom

 

[djm77]

0.03mm finish seems too small. Is the roughing tool causing problems for some reason, meaning that 0.03 for cleaning up isn’t enough. Try leaving 0.2 and see if you get the same issue.


Sent from my iPhone using Tapatalk

 

[metronorth]

ifixcnc;

Mechanically included ballscrew play x and y (none)spindle bearings (good) and whatever else they could think of that was measurable in the machine. They also checked backlash compensation. The ball bar test showed several issues that needed correcting. They were corrected but to no avail.

Hu;

The machine will turn a fine surface finish when facing or cylindrical turning, including some hard turning tests that we recently did.

TDegenhart;

I can't tell you exactly when or how rapidly this started to happen because the gent running it for the last few years made some program changes that masked the problem but slowed the process down. We are trying to get back to where we were originally and that's when we noticed the problem.

To All;

I have done a bit more testing to try and isolate the problem. The "steps" in the surface are actually a scroll, not "steps". I turned the coolant off and ran the finish pass with no coolant so that I could see what was happening and the entire surface is being cut by the finish pass. The pitch of the scroll is far more than the 0.03 mm /rev feed rate that we are using. I turned the feed down to 0.01 mm / rev and achieved an acceptable surface but that obviously is not the proper solution. The challenge continues.

Thanks all, Glenn at Metro North.

 

[CNC Hacker]

Here's a dumb question, are you using the same tool holder for the last 12 years? Possible that the insert seating area is worn out and the insert is moving around under load. What type of tool are you using?

 

[HuFlungDung]

Have a vibration analysis done on the machine. I had an issue caused by an out of balance cooling fan sitting on a 20hp DC motor, on a 3 ton lathe. You wouldn't think a wee vibration all the way back there would show up on the part, but it would. With a the sensor pickup on the vibration analyser, you can probe many different places on the machine and find out that many different vibrations exist. The vibration pattern shown on your part looks quite uniform, as in the sense of a constant vibration affecting the part.

Do you run this with CSS on? It may be up against the rpm limit, I suppose, but if not, then the pitch of the step could show an increase in width if an external vibration is constant.

 

[TDegenhart]

ifixcnc;

Mechanically included ballscrew play x and y (none)spindle bearings (good) and whatever else they could think of that was measurable in the machine. They also checked backlash compensation. The ball bar test showed several issues that needed correcting. They were corrected but to no avail.

Hu;

The machine will turn a fine surface finish when facing or cylindrical turning, including some hard turning tests that we recently did. How about when turning a 45* taper?

TDegenhart;

I can't tell you exactly when or how rapidly this started to happen because the gent running it for the last few years made some program changes that masked the problem but slowed the process down. We are trying to get back to where we were originally and that's when we noticed the problem. I was wondering about a program change but assumed that would be the first place to check.

To All;

I have done a bit more testing to try and isolate the problem. The "steps" in the surface are actually a scroll, not "steps". I turned the coolant off and ran the finish pass with no coolant so that I could see what was happening and the entire surface is being cut by the finish pass. The pitch of the scroll is far more than the 0.03 mm /rev feed rate that we are using. I turned the feed down to 0.01 mm / rev and achieved an acceptable surface but that obviously is not the proper solution. The challenge continues.

Thanks all, Glenn at Metro North.

Tom ....need some characters

 

[Ox]

To All;

I have done a bit more testing to try and isolate the problem. The "steps" in the surface are actually a scroll, not "steps". I turned the coolant off and ran the finish pass with no coolant so that I could see what was happening and the entire surface is being cut by the finish pass. The pitch of the scroll is far more than the 0.03 mm /rev feed rate that we are using. I turned the feed down to 0.01 mm / rev and achieved an acceptable surface but that obviously is not the proper solution. The challenge continues.

Thanks all, Glenn at Metro North.

As was recently asked - you say that you are running CSS, but with only a 2500 RPM limit, I am guessing that this whole feature is ran at 2500? Or is it slowing?

If it is slowing, are you seeing any change in the "scroll" to coincide with the speed change?
Either in distance, or depth?

If not - maybe you could edit your CSS to a level that will actually slow down during the cut and maybe then you could see if there are any changes?


Doo you have any bar whip?
Are you running a 12' feeder?
4' loader?

Any change from the first part in the bar to the last?

If a 12' loader with remnant retract, when near the end of the bar, maybe pull the pusher collet off the bar and run it w/o that touching it. Maybe the pusher collet bearing could be shot?


Considering your feedrate and the scroll pitch not matching, it appears to be a harmonics issue somewhere.


Any change from cold start to end of day? (machine temps)


---------------------

Think Snow Eh!
Ox

 

[adama]

Seeing as its small i would ditch CSS find a smooth rpm and just use it. See if that solves your problem to start with. me, i just plane dont see the point in smaller dia work were theres no tool induced speed limit. You mention ball bar test, but have the axis been checked properly and manually? Got the classic hall marks of stick slip going on there.