Six key indicators to judge the quality of a Datong Okuma vertical machinin

• 1
  Geometric accuracy is the foundation you cannot retrofit
• 2
  Spindle performance decides surface finish and tool life
• 3
  Thermal drift separates a good machine from a lucky one
• 4
  Feed accuracy and repeatability under real load
• 5
  Spindle taper condition and tool holder contact
• 6
  Cutting performance on a known test part
• 7
  Six indicators at a glance: what to measure and what to accept

Every other indicator on this page sits on top of geometry. If the column is not square to the table, no amount of thermal compensation will save a bore that has to line up with a mating plate 300 mm away. Squareness errors do not show up on a single small part. They show up when you flip a part, re-datum it, and the second face is out by 0.03 mm.

We check squareness between X and Y first, then between the spindle axis and the table surface. A working number for a machine in this class is 0.010 mm per 300 mm or better on X–Y, and 0.015 mm per 300 mm on spindle-to-table. Below that, you are paying for temperature control you may not need.

The mistake is measuring geometry on a cold machine and calling it done. Cast iron and steel grow about 11–12 μm per meter per °C. A 20 °C swing across a shop day moves a 1 m column by roughly 0.22 mm if the machine is not compensated. So geometry has to be read together with the thermal behavior, not instead of it.

• 1
  Check firstX–Y squareness, spindle-to-table squareness, then parallelism of the axes.
• 2
  Typical target0.010 mm per 300 mm on X–Y for a machine in this class.
• 3
  Common errorRecording geometry only at cold start and ignoring the afternoon.

The spindle is where the cutting happens, so it sets the ceiling on finish. Two numbers matter more than the catalog rpm: axial and radial runout at the taper, and how the spindle behaves after 30 minutes of cutting. A spindle with 2 μm radial runout will still cut, but you will see the runout in the wall finish and in tool wear on small-diameter end mills.

Ask for a runout reading taken at the gauge line, not at the tool holder flange. A 0.002 mm reading at the flange can become 0.010 mm at 100 mm of tool overhang. That lever arm is why a long reach tool finishes worse than a stub tool on the same machine, even when the spindle is perfect.

Thermal growth of the spindle is the second half of the story. Spindles grow 20–50 μm along Z from a cold start to steady state, and most of that happens in the first 20–40 minutes. If a shop does not warm up the spindle before the first tight part, the first ten parts will be short and the rest will be long. The fix is a warm-up cycle, not a new spindle.

• 1
  Read runout at the gauge lineFlange readings hide the effect of tool overhang.
• 2
  Warm up 20–40 minutesMost Z growth happens in the first half hour of cutting.
• 3
  Watch small toolsRunout hurts a Ø6 mm end mill far more than a Ø50 mm face mill.

Thermal drift is the change in tool position caused by heat, not by wear. Ball screws grow as they run, the bed warms under the work, and coolant pulls heat out of one area while leaving another alone. On a machine without compensation, a 1 m ball screw can move 15–25 μm over a shift. That is enough to fail a ±0.02 mm height callout on a long run.

There are three ways to deal with it. You can run the machine in a temperature-controlled room, you can compensate in the control, or you can accept the drift and re-datum between operations. The first is the most expensive and the most reliable. The second works if the machine has sensors on the screw and the bed. The third is what most job shops actually do, and it is fine if the parts allow it.

We judge thermal behavior with a simple test: cut a test part at cold start, cut the same part after two hours of continuous running, and compare. A machine that holds 0.010 mm between the two is doing well. A machine that moves 0.040 mm needs compensation before it can run tight work unattended.

• 1
  Test methodSame part at cold start and again after two hours of cutting.
• 2
  Good resultWithin 0.010 mm between the two parts.
• 3
  Problem result0.040 mm or more, which needs compensation before unattended runs.

A machine can be accurate when it moves slowly with no load and still miss the mark when it takes a 5 mm depth of cut in 4140 steel. That is why we look at positioning accuracy and repeatability separately. Accuracy is how close the machine gets to the commanded point. Repeatability is how close it comes back to the same point. Repeatability matters more for production because it drives the spread across a batch.

A working target for repeatability on a machine in this class is 0.005 mm on a single axis, measured with a ballbar or a laser interferometer. Below that, you can run tight work without checking every part. Above it, you will be measuring parts more often than you want to.

Backlash is the other half of feed behavior. A worn ball screw or a loose thrust bearing shows up as lost motion when the axis reverses. It hides in circular interpolation tests, where the circle becomes slightly egg-shaped at the quadrant points. If a machine cuts a clean circle in a ballbar test, the feed system is in good shape.

• 1
  Accuracy vs repeatabilityRepeatability drives batch spread, so weight it higher for production.
• 2
  Target0.005 mm repeatability on a single axis.
• 3
  Backlash testCircular interpolation; look for egging at the quadrant points.

Taper contact is easy to ignore and expensive to get wrong. A BT40 or HSK-A63 taper that only touches on 60 percent of its surface will flex under load, and the tool will chatter at speeds where it should be smooth. The fix is not a new spindle. It is checking contact with bluing and regrinding or replacing the taper if the pattern is poor.

We look for 80 percent or better contact on a blued taper, spread evenly around the cone. A patchy pattern near the small end usually means the holder is worn. A pattern that is heavy at the large end means the spindle taper has been damaged, often by a chip that was not wiped off before a tool change.

This indicator connects to the first four. Poor taper contact adds runout, adds heat, and reduces the depth of cut you can take before chatter. On a machine that is otherwise good, fixing the taper can recover most of the performance you thought was lost to age.

• 1
  Target contact80 percent or better, evenly spread around the cone.
• 2
  Patchy small endUsually a worn tool holder, not the spindle.
• 3
  Heavy large endSpindle taper damage, often from a chip at tool change.

The last indicator is the one that ties everything together: cut a part you already know. We use a test piece with a bored hole, a faced surface, and a tapped hole at a known position. The bore checks roundness and size. The face checks flatness and finish. The tapped hole checks position and thread quality.

Run the part in a material the machine will actually see. Aluminium 6061 cuts differently from 17-4PH stainless, and a machine that holds ±0.005 mm in aluminium may hold ±0.015 mm in stainless because of cutting forces and heat. Match the test to the job, or the result will not mean much.

Compare the finished part to the drawing, not to the machine display. The display shows where the machine thinks it is. The part shows where it actually went. When those two disagree, the machine has a problem that no amount of parameter tuning will hide.

• 1
  Test featuresBored hole, faced surface, tapped hole at a known position.
• 2
  Match the materialAluminium and stainless give very different results on the same machine.
• 3
  Trust the partThe display shows intent; the finished part shows reality.