To avoid confusion, we first define each term with plain-language framing, rooted in the ISO 230-2 standard that governs CNC machine performance testing.
Accuracy is the degree of alignment between a machine’s intended position (the target value programmed into the CNC controller) and its actual physical position. Think of it as "how close the machine gets to where it’s supposed to be." For example, if you program a CNC machine to drill a hole at X=100.000 mm, Y=50.000 mm, accuracy describes the difference between the programmed target and the hole’s real-world location.
Repeatability is the degree of consistency of a machine’s actual position when it repeatedly returns to the same programmed target. It measures "how often the machine lands in the same spot when asked to go there multiple times." Using the same drilling example, if you program the machine to drill 10 holes at X=100.000 mm, Y=50.000 mm, repeatability describes the variation between the 10 holes’ actual locations.
A simple analogy clarifies the difference: A archer aiming at a target’s bullseye has:
For CNC machines, these values are specified per axis (X, Y, Z, and rotary axes for multi-axis models) and follow ISO 230-2 testing protocols, which require multiple bidirectional runs (the machine approaches the target from both positive and negative directions) to capture real-world performance.
The distinction between accuracy and repeatability directly impacts part function, assembly compatibility, and production cost.
Parts with critical geometric tolerances depend on both metrics. For example:
In modular assembly, parts from different production runs must fit together seamlessly. Poor repeatability leads to inconsistent part dimensions, requiring manual rework or scrapping. Poor accuracy leads to systematic misalignment across all parts from a machine, requiring fixture adjustments or design modifications.
CNC machine manufacturers and production teams use standardized methods to measure accuracy and repeatability, with inherent limits to each method.
The gold standard for measuring both metrics, laser interferometry uses a laser beam to track the machine’s axis position with sub-micron precision. The machine moves along an axis to a series of programmed targets, and the laser records the actual position for each run.
Practical shop-floor note: Laser interferometry requires a controlled environment (stable temperature, minimal vibration) and trained personnel, making it a periodic test (typically annual or after machine relocation) rather than a daily check.
Ball bar testing measures circular interpolation accuracy and repeatability, critical for machining curved features. A precision ball bar is mounted between the machine’s spindle and table, and the machine moves in a circular path. Deviations from the ideal circle are recorded to calculate:
Practical shop-floor note: Ball bar testing is faster than laser interferometry and can be performed by trained operators, making it a common quarterly check for production machines.
Modern CNC machines use touch probes mounted in the spindle to measure part dimensions during machining. Probing can verify both accuracy (comparing a measured feature to the programmed target) and repeatability (measuring the same feature multiple times).
Practical shop-floor note: Probing accuracy is limited by probe tip wear and machine thermal drift, so it is best used for relative measurements rather than absolute calibration.
Manufacturers achieve specified accuracy and repeatability through design choices:
For example, Gree CNC’s GA-DV750 vertical machining center has a specified positioning accuracy of 0.003 mm and repeatability of 0.0024 mm, achieved through high-precision linear guides and a closed-loop feedback system.
Accuracy and repeatability are degraded by different sources of error, explaining why the two metrics differ for all CNC machines.
Accuracy errors are systematic, meaning they affect all runs in the same direction. Common sources include:
Repeatability errors are random, meaning they vary between runs. Common sources include:
For all CNC machines, repeatability is always better (lower value) than accuracy. This is because accuracy includes both systematic and random errors, while repeatability only includes random errors. For example, Gree CNC’s GA-MV856 has a positioning accuracy of 0.010 mm and repeatability of 0.006 mm: the 0.004 mm difference comes from systematic errors (e.g., thermal drift, calibration offset) that affect all runs consistently.
It is critical to understand what accuracy and repeatability measurements do not tell you, and common overconfidence in these metrics.
Many production teams assume that a machine’s specified accuracy and repeatability will be maintained indefinitely. In reality, these metrics degrade over time due to wear, thermal cycling, and environmental changes. For example, a machine with an initial repeatability of 0.004 mm may degrade to 0.01 mm after 2 years of heavy use without maintenance.
| Measurement Method | Accuracy Measurement Limit | Repeatability Measurement Limit | Effort | Typical Use Case |
|---|---|---|---|---|
| Laser Interferometry | ±0.001 mm | ±0.0005 mm | High (specialized equipment, trained personnel, controlled environment) | Annual machine calibration, after relocation or major maintenance |
| Ball Bar Testing | ±0.002 mm | ±0.001 mm | Medium (trained operator, 1-2 hours per test) | Quarterly production machine performance checks |
| In-Process Probing | ±0.005 mm | ±0.003 mm | Low (automated, operator-initiated) | Daily part quality verification, in-process adjustments |
Use this checklist to apply accuracy and repeatability correctly in production:
Can a machine have high repeatability but low accuracy? Yes. This occurs when there is a systematic offset in the machine’s position. For example, a machine may consistently drill holes 0.01 mm to the left of the programmed target (low accuracy) but all holes are within 0.002 mm of each other (high repeatability). This is common and can be corrected with a one-time calibration adjustment.
Which is more important for production? It depends on the part. For parts requiring alignment with mating components (e.g., engine blocks, medical devices), accuracy is more important. For high-volume parts requiring consistent dimensions (e.g., gears, fasteners), repeatability is more important. In most cases, both are critical for overall production quality.
How often should I test accuracy and repeatability? Laser interferometry (gold standard) should be performed annually, after machine relocation, or after major maintenance. Ball bar testing should be performed quarterly to monitor performance between full calibrations. In-process probing checks should be performed daily to verify part quality.
Do multi-axis machines have different accuracy and repeatability? Yes. Multi-axis machines (e.g., 5-axis) have additional rotary axes, each with their own accuracy and repeatability. Combined movement of linear and rotary axes can lead to lower overall accuracy than individual axis measurements suggest. For example, Gree CNC’s GA-FA320 5-axis machining center has a specified accuracy of 0.010 mm and repeatability of 0.006 mm, which applies to combined axis movement.
Can I improve a machine’s accuracy and repeatability after purchase? Yes. Accuracy can be improved through calibration, thermal compensation adjustments, and fixture alignment. Repeatability can be improved through preventive maintenance (e.g., replacing worn ball screws, adjusting spindle runout) and environmental control (e.g., temperature stabilization). However, there are limits to how much performance can be improved; a machine’s design sets the maximum achievable accuracy and repeatability.
Send us your part, batch size and target port. We reply with a configuration, a binding lead window and a total-cost comparison.
Request a quote →