Mill Accuracy

[Downwindtracker2]

My mill is a RongFu-45, a square column mill/drill. It had been CNCed and was well worn when I started working on it. On your machines what kind of accuracy do you expect ?

 

[Megar arc 5040dd]

I have a clone of the RF-45 that I picked up used in really good shape. The manufacture date on it is from 07 if I recall. I expect to be within .0005" and so far have been able to achieve that about 50% of the time. The rest of the time I am within .001". I have no reason yet to believe the problem is the machine I am 99% sure I am the accuracy issue. The machine could be more accurate but I have no way of knowing because I'm not good enough to achieve that type of accuracy.

Long story short my Chinese knock off is more accurate than I am and more that accurate enough for hobby use.

What type of accuracy do you need or want?

 

[slow-poke]

I use a closed loop servo with the inner loop closed around the linear scales not the encoder on the servo. Closing the loop this way eliminates the errors in the ball screws, bearings etc. tolerance is set to 0.00039, however the machine must flex to some extent so who knows how much and under what load?

I do know it's a lot better (probably 10x better) than the open loop stepper system on my now sold old round column mill.

 

[Dabbler]

I use a manual mill with a 2 tenths DRO on it. After making some plates my work has measured to just over 2 tenths over a 30" X 30" aluminum die plate. I had to machine at 23C, as the use environment is 23C. One degree C works out nearly a thou difference between holes at the extreme edges.

It took forever to get these plates right, especially as the work envelope was 19"X13", and they had to be repositioned 3 times to finish the job.

 

[LenVW]

Remember to use your ‘axis locks’ to reduce slack that will increase as your mill ages.
Whether you have HAAS, TOSHIBA or off-shore workcenters there will be a degree of play in the leadscrews. The key is to use the correct practises to compensate for the play between components.
Also,
Always feed your cutters in the direction of ‘fixed jaws’ and secure locating surfaces.

 

[Chipper5783]

As stated by others, the dominant factor in accuracy is the operator. I’m pretty sure your machine is more capable than anything available in the year 1800, and a good hand back then could do amazingly accurate work. Really the issue is speed and having to compensate for short comings - today’s “accurate” machine should need very little compensation - set your numbers and perfect results (of course reality is not like that, there will nearly always be some consideration due to less than perfect in the machine, the setup, the tooling etc). That is one of the reasons that identifying target tolerances is important: sometimes an acceptable part has to be +/- 0.0001”, sometimes it is +/-0.01” in which case achieving said result involves different approaches.

 

[Dabbler]

Chipper is 100% - even in manual machining, you leave stock allowance for handwork to get the ultimate accuracy... Always dependent on the required tolerance.

I knew a toolmaker that would refuse to make anything with greater than 4 tenths tolerance, and he was slow, SLOW. He was unnecessarily spending time not appreciated by the user of the build... His custom triggers wouldn't work in the cold, because his tolerances were too tight, and you couldn't recock the trigger in -10C weather. sigh.

 

[Tom Kitta]

I use a manual mill with a 2 tenths DRO on it. After making some plates my work has measured to just over 2 tenths over a 30" X 30" aluminum die plate. I had to machine at 23C, as the use environment is 23C. One degree C works out nearly a thou difference between holes at the extreme edges.

It took forever to get these plates right, especially as the work envelope was 19"X13", and they had to be repositioned 3 times to finish the job.

This is more or less exactly like my mill - I was shocked when I got within 1 thou on 14 in - I did not know my vise was that well positioned.