Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
ASME B89.4.1-1997 — Methods for Performance Evaluation of Coordinate Measuring Machines
📐 ASME B89.4.1-1997, Methods for Performance Evaluation of Coordinate Measuring Machines, is the foundational document for coordinate measuring technology within the ASME B89 dimensional metrology standards family. It does not attempt to attach a single “accuracy grade” to a coordinate measuring machine (CMM) or to dictate a unique pass/fail criterion. Instead, it establishes a reproducible, comparable, traceable testing language: a uniform terminology, standardized test layouts, and a structured specification-form format that fully describe a machine’s behavior at four levels — the environment, individual geometric performance, volumetric performance, and the probing subsystem. The 1997 edition was revised by two addenda (1998 and 2001); Addenda b-2001, approved by the American National Standards Institute (ANSI) on August 31, 2001, was issued as replacement pages incorporated directly into the text, with substantive revisions to Section 1, Section 5.5 on volumetric performance, and Section 6 on sub-system performance tests.
💡 The key to understanding B89.4.1 is its “bottom-up” philosophy: decompose the machine into 21 geometric error components and several subsystem capabilities, measure and report each one, and let the user decide whether the machine is adequate for his or her own tolerances. This complements the “overall length measurement error” approach later adopted by ISO 10360-2.
1. Purpose and Scope 📌
The Standard was developed by Division B89.4 (Coordinate Measuring Technology) of the ASME B89 Standards Committee on Dimensional Metrology. Its seven chapters reflect a clear engineering logic: Chapter 1 Scope; Chapter 2 Definitions; Chapter 3 Environmental Specifications; Chapter 4 Environmental Tests; Chapter 5 Machine Performance; Chapter 6 Sub-System Performance Tests; and Chapter 7 Test Equipment. The structure itself embodies the sequence — first confirm that the measurement environment is adequate, then evaluate overall machine performance, and finally probe into the probing subsystem.
The Standard applies to Cartesian-type coordinate measuring machines of many architectures, including moving bridge, fixed bridge, fixed table cantilever, column, horizontal arm (moving ram and moving table), L-shaped bridge, and gantry machines, as well as compound machines with rotary tables. For large dual-ram machines sharing a common base and operating in the duplex mode, dedicated test arrangements are also provided.
A point that deserves emphasis is the Standard’s “method” character: it prescribes how to measure, how to record, and how to state results, not what performance must be achieved. Acceptance limits are agreed between buyer and seller in the contract; B89.4.1 provides the common basis for writing those limits down clearly and measuring them correctly.
⚠️ The Standard does not apply to non-Cartesian measuring systems such as articulated arms, laser trackers, or photogrammetric systems. Their evaluation is covered by B89.4.19, B89.4.22, and B89.4.23; mixing test methods makes performance data incomparable.
2. Terminology and Machine Classification 🔧
Chapter 2 builds an extensive glossary whose concepts are prerequisites for understanding every test:
- Abbe offset and Abbe error: when the axis of the measured length and the axis of the machine scale are not collinear, any angular motion is amplified into a reading error by the offset distance — roughly the product of offset and angular deviation. The Standard illustrates this mechanism schematically; it is the fundamental reason the same machine can yield different results for the same gauge block measured at different locations.
- Hysteresis: the difference in readings when the probe approaches the same target from opposite directions, reflecting the combined effects of drive backlash, friction, and preload.
- Repeatability: the scatter of repeated measurements of the same length under identical conditions — the signature of random error.
- Volumetric performance: measurement of calibrated lengths in many directions and positions throughout the working volume, giving an integrated view of spatial indication consistency.
- Vectoring performance: the machine’s ability to move along a commanded direction while maintaining probing capability, involving dynamic and servo behavior.
| Machine Type | Structural Feature | Typical Application | Evaluation Focus |
|---|---|---|---|
| Moving bridge | Bridge travels in Y; carriage travels on the beam | Batch inspection of small/medium parts | Abbe errors from bridge pitch and beam straightness |
| Fixed bridge | Bridge fixed; table moves | High-accuracy small/medium parts | Table angular motions and load deformation |
| Fixed table cantilever | Column cantilevered; table fixed | Large, heavy workpieces | Cantilever sag and Y-axis angular deviation |
| Column | Vertical column drives probe lift | Box-type parts | Z-axis straightness and counterweight effects |
| Horizontal arm | Ram extends horizontally; moving-ram or moving-table layout | Automotive body-in-white, large panels | Ram extension sag and angular deviation |
| L-shaped bridge | Bridge forms an L to enlarge range | Wide workpieces | L-structure torsion and thermal symmetry |
| Gantry | Two columns with large span; synchronized beam drive | Very large workpieces, dies | Dual-drive sync errors, foundation and thermal gradients |
💡 The same test method applied to different architectures produces completely different error maps. When selecting test items, prioritize the “weak degrees of freedom” of the structure — for a horizontal-arm machine, examine the influence of ram extension on indications; for a gantry, examine structural bending under thermal gradients.
3. Technical Requirements and Core Tests 📏
Chapter 5 (Machine Performance) and Chapter 6 (Sub-System Performance Tests) form the technical core. The former addresses the machine as a whole, the latter the probing system. The purpose and method of each test are summarized below:
| Test | Object Evaluated | Method Outline | Output |
|---|---|---|---|
| Hysteresis | Drive backlash, guide friction | Approach the same target from both directions along each axis and read | Forward/reverse reading difference |
| Repeatability | Random error level | Repeat the same length measurement many times at a fixed position and probe | Range or standard deviation of the series |
| Linear displacement accuracy | Single-axis scale-to-truth consistency | Compare axis position indications with gauge blocks, step gauges, or a laser interferometer | Positional indication errors |
| Volumetric performance | Spatial indication consistency | Measure calibrated lengths in many directions and positions in the volume | Length error distribution by direction |
| Bidirectional length measurement capability | Forward/reverse measurement consistency | Complete measurements of the same length in both directions | Bidirectional result difference |
| Probing — point-to-point | Single-touch probing repeatability and offset | Repeatedly trigger on the same target point of a sphere or gauge | Probe repeatability and pre-travel scatter |
| Probing — multiple-tip | Consistency among probe/stylus combinations | Measure the same feature with different probe configurations and compare | Inter-probe consistency deviation |
| Probing — scanning | Continuous scanning filtering and following | Scan a master part at specified speed and compare profiles | Scan indication error vs. speed |
| Vectoring performance | Motion along the commanded direction | Move along the commanded direction and trigger probing | Direction-keeping deviation |
Another hallmark of the Standard is its Specification Form system. Figures 1A, 1B, and 1C give the standard formats for the machine specification form, the environmental specification form, and the performance specification form. When the manufacturer completes these three forms, the buyer obtains a structured, comparable performance declaration at the procurement stage — no reliance on vague marketing statements.
✅ Recommended practice: attach the three B89.4.1 specification forms directly to the procurement technical agreement, and require the supplier to complete every field together with the test conditions (probe model, stylus length, workpiece load, environmental temperature). This single measure is often more effective than repeated negotiation of accuracy numbers.
4. Environmental Specifications and Tests 🌡️
Chapter 3 brings the environment itself under specification management, requiring four categories of parameters to be stated explicitly:
- Temperature: not only a nominal value (normally referenced to 20 °C) but also spatial gradient, rate of change over time, and short-term fluctuation. For steel parts, a 1 °C deviation produces roughly 11.5 μm/m of length error — temperature is the dominant error source in most field measurements.
- Vibration: permitted amplitude and frequency spectrum; the relative-motion test (Section 4.3) verifies machine stability under actual foundation and ambient vibration sources.
- Electrical: allowable voltage, frequency, harmonic distortion, grounding, and transient interference.
- Utility air: pressure, dew point, oil content, and particulate level for air bearings and pneumatic components.
Chapter 4 provides the corresponding test methods. The thermal test is of greatest practical value: by applying a temperature step or simulating a typical operating cycle and recording the response of critical machine locations, one obtains the thermal time constant and steady-state drift. These data directly determine how long a machine must warm up before precision measurement can begin, and whether a temperature-controlled room or a local enclosure is needed.
🚨 A common mistake: performing acceptance tests directly under an HVAC diffuser, in sunlight, or next to a busy aisle. Such layouts push gradient and fluctuation beyond specification, so the measured volumetric data neither represents the machine’s true capability nor can be reproduced on reverification. Stabilize for at least 24 hours and record temperature continuously throughout the test.
5. Engineering Practice Essentials 🛠️
- 🔍 Three test types serve different purposes: acceptance tests (new machine delivery, most complete), reverification tests (periodic confirmation that the machine still meets its original specification — key items only), and interim checks (trend monitoring between reverifications, typically with a reference sphere or uni-directional artifact, low cost and high frequency).
- 🔧 Probe and stylus conditions must be recorded: the same machine with different stylus lengths and masses behaves differently in dynamic terms. Performance data without the stylus configuration noted are effectively incomparable.
- 📊 Workpiece loading effects: the Standard requires consideration of workpiece weight on table and structure deformation. Heavy parts should be measured under load, or corrected using the manufacturer’s load-deformation curves.
- 📋 Uncertainty statements: results should carry measurement uncertainty statements and note the traceability chain of the standards used — systematized later in B89.7.1 and B89.7.5.
- 🧭 Interface with ISO 10360-2: if the customer or supply chain requires metrics such as E0, E150, or R0, test in accordance with B89.4.10360.2; they cannot be derived from B89.4.1 data.
💡 Direction selection in volumetric testing matters: the three axis directions expose single-axis cumulative error; the three face diagonals are most sensitive to two-axis linkage angular errors; the body diagonal combines all 21 error components. A systematic discrepancy among the three groups often indicates a squareness problem rather than a mere scale error.
6. Frequently Asked Questions ❓
❓ Q: Has B89.4.1-1997 been superseded by the ISO 10360 series?
A: No. B89.4.1 follows a component-by-component evaluation route that can localize error sources; ISO 10360-2 (ASME B89.4.10360.2) follows an overall-performance route that makes conformity decisions easy for buyer and seller. Common engineering practice: acceptance by overall metrics, diagnostics and accuracy improvement by component tests.
A: No. B89.4.1 follows a component-by-component evaluation route that can localize error sources; ISO 10360-2 (ASME B89.4.10360.2) follows an overall-performance route that makes conformity decisions easy for buyer and seller. Common engineering practice: acceptance by overall metrics, diagnostics and accuracy improvement by component tests.
❓ Q: Why is hysteresis measured separately — isn’t it the same as repeatability?
A: The mechanisms differ. Repeatability describes the scatter of repeated measurements in the same direction (random error); hysteresis describes the systematic difference when approaching the same target from opposite directions, caused by backlash, friction, and preload. A machine with large hysteresis shows a “return difference” in scanning or bidirectional work that averaging cannot remove.
A: The mechanisms differ. Repeatability describes the scatter of repeated measurements in the same direction (random error); hysteresis describes the systematic difference when approaching the same target from opposite directions, caused by backlash, friction, and preload. A machine with large hysteresis shows a “return difference” in scanning or bidirectional work that averaging cannot remove.
❓ Q: What did Addenda b-2001 mainly change?
A: b-2001 was the second addenda to the 1997 edition. Besides adding a “Correspondence With the B89 Committee” section and updating the contents, it revised Section 5.5 on volumetric performance (adding subcategories and clarifications), rewrote 5.6.2 and 5.6.3 on bidirectional length measurement capability, and expanded Chapter 6 sub-system performance tests with new Tables 4–7 and new Sections 6.3 and 6.4. When using an old single-volume copy, verify that the addenda replacement pages have been incorporated.
A: b-2001 was the second addenda to the 1997 edition. Besides adding a “Correspondence With the B89 Committee” section and updating the contents, it revised Section 5.5 on volumetric performance (adding subcategories and clarifications), rewrote 5.6.2 and 5.6.3 on bidirectional length measurement capability, and expanded Chapter 6 sub-system performance tests with new Tables 4–7 and new Sections 6.3 and 6.4. When using an old single-volume copy, verify that the addenda replacement pages have been incorporated.
❓ Q: Can a small company without a laser interferometer run only part of the tests?
A: Yes. Section 1.2 (Alternatives) explicitly permits equivalent methods by mutual agreement, such as calibrated gauge blocks or step gauges instead of an interferometer for linear displacement and volumetric tests. The method must be documented, and the uncertainty of the reference standards should be significantly smaller than the machine’s permitted error (typically better than 1:4).
A: Yes. Section 1.2 (Alternatives) explicitly permits equivalent methods by mutual agreement, such as calibrated gauge blocks or step gauges instead of an interferometer for linear displacement and volumetric tests. The method must be documented, and the uncertainty of the reference standards should be significantly smaller than the machine’s permitted error (typically better than 1:4).
