Views: 0 Author: Fannie Chen Publish Time: 2026-08-19 Origin: SZGH
Verifying CNC machine accuracy before shipment comes down to three independent checks: a ball-bar test to catch geometric and servo errors during circular motion, a laser interferometer test to measure positioning accuracy and repeatability on each individual axis under ISO 230-2, and a sample-cut evaluation to confirm the machine performs correctly under real cutting loads — not just running empty. A supplier's stated tolerance figure means little without knowing which of these tests produced it, under what conditions, and with a report you can actually read. This article explains what each test measures, how it's actually performed, and how to interpret the pre-shipment inspection report a supplier should be able to provide.
Positioning accuracy and repeatability are two different numbers, not interchangeable terms — a machine can have excellent repeatability with mediocre positioning accuracy, or vice versa.
A ball-bar test detects geometric and servo problems (backlash, squareness errors, servo mismatch) in minutes by tracing a circular path and comparing it to the programmed circle.
A laser interferometer test measures positioning accuracy and repeatability on each linear axis individually, per the ISO 230-2 standard, and is the definitive method for geometric accuracy verification.
A sample-cut test reveals problems that geometric tests alone can miss — thermal drift, chip evacuation issues, vibration under actual cutting load — because it's the only test performed with the machine actually cutting material.
A legitimate pre-shipment inspection report names the test standard used, the specific equipment (not just "measurement equipment"), the measured values with units, and the conditions under which the test was run — a report without these details isn't independently verifiable.
Positioning accuracy and repeatability answer two different questions, and a supplier quoting only one of them — or blending the two into a single "precision" claim — hasn't given you the full picture.
Positioning accuracy measures how close the machine actually gets to a commanded position — the maximum deviation between where the machine was told to go and where it actually ended up. This is measured across the full travel range, typically with bidirectional readings (approaching the target from both directions) at evenly spaced intervals.
Repeatability measures how consistently the machine returns to the same commanded position across multiple attempts — it doesn't tell you whether that position is correct, only whether the machine lands in the same place every time.
A machine can have excellent repeatability (it consistently returns to the same spot) while that spot is still measurably off from the commanded position — meaning positioning accuracy is worse than repeatability suggests. This is why both numbers need to be reported separately, and why a single "precision" figure without specifying which one it refers to should prompt a follow-up question. Both are defined and measured under ISO 230-2, the international standard for testing positioning accuracy and repeatability of CNC machine axes, using repeated measurements at each tested position.
A ball-bar test measures a CNC machine's positioning performance during circular motion — if a machine is instructed to trace a perfect circle and its axes are perfectly synchronized and free of geometric error, the resulting path matches the programmed circle exactly; in practice, deviations reveal specific, diagnosable problems.
The test procedure works like this:
A telescoping ballbar device — a precision linear sensor with a magnetic ball at each end — is mounted between the machine's table and spindle (or tool holder), using kinematic magnetic joints that let it pivot freely while measuring length changes.
A simple program commands the machine to trace a circular path, typically in both clockwise and counterclockwise directions, with a small extra arc added before and after to account for the machine's acceleration and deceleration.
As the machine moves, the ballbar continuously measures tiny variations in the radius between the two mounting points, capturing data in real time.
Software compares the actual traced path to the ideal programmed circle and calculates deviation metrics.
A ball-bar test typically takes only a few minutes to run, but reveals a specific set of problems that are otherwise difficult to isolate: servo mismatch and lag (one axis responding faster or slower than another), backlash and lost motion (play in the mechanical drive system), reversal spikes (a brief glitch when an axis changes direction), and squareness errors between axes (whether X and Y, for example, are truly perpendicular). Larger test radii are more sensitive to overall machine geometry errors, while smaller radii are more sensitive to servo mismatch and lag — which is why test radius should be selected relative to the specific concern being diagnosed.
A laser interferometer test measures an individual axis's positioning accuracy and repeatability directly, with sub-micron resolution, and is the definitive method for geometric accuracy verification under ISO 230-2 — where a ball-bar test gives a fast overall health check, a laser interferometer test gives the detailed, per-axis numbers that a formal acceptance report is actually built on.
The test procedure, at a general level:
Before testing begins, the machine and measurement environment are stabilized — this includes allowing the machine to reach thermal equilibrium (often through a warm-up run) and controlling ambient temperature, since thermal expansion directly affects measured position.
The laser interferometer is set up in the same plane as the axis being tested, typically using a laser head, a reflector mounted where the tool or table would be, and an optical path aligned with the axis of travel.
The machine moves the axis to a series of target positions across its full travel range. At each position, the laser measures the actual distance traveled using interference patterns created by splitting and recombining the laser beam — a method precise enough to resolve positioning errors at the micron level.
Measurements are typically taken bidirectionally (approaching each target position from both directions) and repeated multiple times at each position, which is what allows both positioning accuracy and repeatability to be calculated from the same data set.
Beyond simple linear positioning, laser interferometer systems can also measure straightness, squareness, and angular positioning errors, providing a fuller picture of the machine's geometric accuracy.
Because this method isolates each axis individually under controlled, repeatable conditions, it's the test result that should appear on a formal, standards-referenced inspection report — a "laser report" citing ISO 230-2 is meaningfully different from a general claim of "verified accuracy" with no named method.
A sample-cut test — running an actual cutting program on real material, not just an empty (no-load) motion test — reveals problems that geometric accuracy tests alone can miss, because it's the only test performed under the thermal load, vibration, and cutting forces the machine will actually experience in production.
Geometric tests like ball-bar and laser interferometer measurements are typically performed with the machine running empty or under controlled, non-cutting conditions. This is by design — it isolates the machine's inherent geometric and servo performance from the additional variables that cutting introduces. But those additional variables are exactly what a buyer ultimately cares about: does the machine hold its stated tolerances once it's actually removing material?
A sample-cut evaluation should, at minimum, confirm:
Dimensional accuracy on a representative part — ideally your own part drawing or a comparable geometry, not just a generic test block, measured against the drawing's actual tolerances.
Surface finish under the specific cutting parameters (speed, feed, tool, material) relevant to your application.
Thermal stability over a realistic cycle — a machine that holds tolerance at the start of a run but drifts as the spindle and structure heat up is a real risk that a short geometric test won't catch.
Chip evacuation and vibration behavior under actual cutting loads, which can affect both finish and tool life in ways a no-load test cannot reveal.
Requesting a sample-cut demonstration — ideally using your own part drawing — before final acceptance is one of the more reliable ways to confirm a machine's real-world performance matches its published specifications.
A pre-shipment inspection report is only as useful as its specificity — a report should name the test standard, the equipment used, the numeric results with units, and the conditions under which testing was performed, not just assert that "accuracy was verified."
At minimum, look for:
The test standard referenced (for example, ISO 230-2 for positioning accuracy and repeatability), not just a general claim of compliance.
The specific equipment used — named as a laser interferometer, ball-bar system, or CMM (coordinate measuring machine), not a vague "measurement equipment" line item.
Separate, labeled values for positioning accuracy and repeatability, with units and the travel length or range they apply to (accuracy specs are typically expressed relative to a specific length, such as per 300mm of travel).
The conditions under which testing was performed — whether the machine was warmed up, what ambient temperature was maintained, and whether the test was static or included dynamic (in-motion) measurement.
Sample-cut results if applicable, including the part measured and the specific dimensions checked against drawing tolerances.
A report that only states a single tolerance figure with no named test method, standard, or equipment doesn't give you anything independently verifiable — you're taking the number on faith rather than confirming it.
Do I need to request all three tests (ball-bar, laser interferometer, sample-cut), or is one enough?
They check different things and aren't substitutes for each other. A ball-bar test is a fast, broad health check; a laser interferometer test gives the detailed per-axis numbers a formal report should be built on; a sample-cut test confirms real-world cutting performance. For a significant purchase, requesting evidence from all three gives a genuinely complete picture rather than a partial one.
Can I request these tests to be performed on my own part drawing rather than a generic test piece?
Yes, and for the sample-cut test specifically, this is generally more informative than a generic test block, since it confirms performance on the actual geometry, material, and tolerances you care about rather than a standardized but potentially unrepresentative shape.
What if a supplier's report only states one overall "precision" number instead of separate positioning accuracy and repeatability figures?
This is worth following up on directly — ask which of the two the number represents, and request the other figure along with the test method and standard used. A single blended number without this detail is harder to independently evaluate.
Is a laser interferometer test something I need to witness in person, or can it be done remotely?
A live video demonstration of the test being performed is a reasonable middle ground for buyers who can't travel — it lets you confirm the test is actually being run on your specific machine rather than reusing a generic report, even if you can't be physically present for the full procedure.
How does this relate to CE certification — doesn't that already confirm the machine is accurate?
No — CE certification addresses safety, electromagnetic compatibility, and compliance with EU machinery directives, not machining accuracy. A CE-certified machine can still have widely varying positioning accuracy and repeatability depending on its build quality; accuracy verification is a separate process from safety/compliance certification.
Verifying CNC machine accuracy before shipment means going beyond a single stated tolerance figure and confirming, with named methods and standards, how that figure was actually produced. A ball-bar test catches geometric and servo problems quickly; a laser interferometer test provides the detailed, standards-referenced positioning accuracy and repeatability figures under ISO 230-2; and a sample-cut test confirms the machine performs under real cutting conditions, not just when running empty. Together, these three checks — documented in a report that names its methods, equipment, and conditions — give you something you can actually verify, rather than simply trust.
Request a Pre-Shipment Inspection Report — Ask SZGH's engineering team for the specific test methods, standards, and documented results for your machine before it ships. Email: export02@szghtech.com · WhatsApp: +86-18925223781
2026-07-23 1124
SZGH-Technology-Full-Product-Catalog-Robots-CNC-Automation-2026.pdf
2026-06-18 19
SZGH CNC Milling Controller Catalog.pdf.pdf
2026-06-17 3
SCARA Robot White Paper.pdf
2026-06-11 21
SZGH-Collaborative-Robot-Cobot-Catalog-BCi-Series.pdf
2026-06-10 64
Shenzhen Guanhong Technology - Servo Motor Brochure 2025.4.pdf
2026-05-11 37
CNC MACHINE TOOL CATALOG.pdf
