Oscillation During Axis Acceleration

[Geof]

Due to the correct machine being out of action we put a part onto our GR510 even though it is not really a suitable machine. The part has a hole and a circular profile that we like to keep within a two thou range but which can be allowed to drift up to twice this; using the gantry was pushing it because Haas claims: Positioning (±) 0.0020 " and Repeatability (±) 0.0010 " .

More than 99% of the parts came off within two thou and the lower flange face which is done in three passes with the tool lifting clear between each pass was smooth to the touch between adjacent toolpaths indicating the machine was holding a better tolerance than spec at least in repeatability of the Z axis. Each load was 16 parts and an interesting thing was that a bad part could be between two good parts.

Also interesting was the feature shown in the picture. The profile is a series of blended radii with a straight front that is not tangential to the curve on either end. In the picture it is possible to see ripples where the X axis stops at the transition from G02 to G01 and then again much more obvious at the G01 to G02 transition where X has to accelerate. This is not really surprising because the X axis is the entire gantry structure which I estimate weighs around 6000lb; obviously something this massive needs a healthy push to stop and start. The pattern of ripples was consistent from part to part and as the picture shows at 60 ipm they are smaller than at 100 ipm.

[JPMach]

We had the same sort of problem at a shop I used to work at. We had a big shinx 4 head router with a new camsoft control upgrade (just about tore my hair out putting that darn thing in). It seemed like the axis would never come to an instant stop even from say 50 ipm, it would bounce back and forth half a dozen times or so before it came to rest and it was consistent doing it too. We tried everything. What worked the best was manually going in and putting slow down points in where it would turn sharp corners and small radiuses. In order to do a straight line cut at 400 ipm I would have to edit in half a dozen 2-3 inch long sections with increasingly higher speeds and then the same at the end to slow down.

Of course the boss got the machine cause it was cheap so I think he learned his lesson after several months on the phone to tech support trying to make it run better and paying me to spend countless hours going back and retuning the servos several times.

Just get the VMC fixed and put them back in there!

JP

[Geof]

Originally Posted by JPMach ....We tried everything. What worked the best was manually going in and putting slow down points in where it would turn sharp corners and small radiuses. In order to do a straight line cut at 400 ipm I would have to edit in half a dozen 2-3 inch long sections with increasingly higher speeds and then the same at the end to slow down......JP

That is an interesting but tedious way to combat the problem. I threw a few G09s which reduced but did not elimnate the problem

It is a pity a command does not exist for controlled acceleration both positive and negative over a specified distance.


Just get the VMC fixed and put them back in there!

(Meekly) Yes Boss.

Is buying a new VF2, SMTC, 15,000 spindle, 1000 ipm rapids acceptable? I got lucky! Haas had an order cancellation and I picked up a lot of options at no cost; sometimes the universe does unfold in the correct way.

[JPMach]

Must be nice to money laying around like that :-)

JP

[Geof]

Originally Posted by JPMach Must be nice to money laying around like that :-) JP

It is but back when I was your age my only significant possession was a fourteen year old Dodge sedan. I was in my first year at college and working afternoon shift three days a week as a machinist.