Surface roughness testers
Selecting surface-texture instruments and reporting roughness with controlled parameters, filtering, trace geometry and uncertainty.
Tags: inspection, surface-texture, roughness
On this page
Specify the parameter, units, location and evaluation settings. A profile or area supplies the heights used in the calculation.
Illustrative arithmetic, not a complete instrument evaluation: four equally weighted, already mean-centred profile heights are −1, −1, +1, +1 µm. Their mean absolute height is 1.0 µm, and RMS height is 1.0 µm. For −2, 0, 0, +2 µm, the mean absolute height is still 1.0 µm, but RMS height is √2 ≈ 1.414 µm. Equal average absolute height therefore does not imply equal distribution of peaks and valleys. Real parameter calculations require the specified filtering and evaluation domain.
Instrument families and references
A contact stylus instrument traverses a tip over the surface and records vertical displacement. A skidded instrument takes its reference partly from a shoe riding on the part; a skidless arrangement uses an instrument datum. These configurations respond differently to longer-wavelength form and waviness. NPL's GPG37, Issue 2, updated February 2014, describes instrument construction, profile measurement and skid references. Use its physical guidance while checking the standard edition invoked by the drawing. NPL, The Measurement of Surface Texture using Stylus Instruments
Optical profilers provide another route, including areal measurement. Avoid assuming a contact profile and an optical area measurement are equivalent because both displays say “roughness”. KEYENCE distinguishes line-based Ra/Rq from areal Sa/Sq and their different sampling domains. Select the required measurand first, then qualify the method for the material and feature. KEYENCE, Ra Surface Roughness vs. Sa, undated technical article
Parameters and drawing interpretation
| Parameter | Interpretation to preserve |
|---|---|
| Ra | Mean absolute profile-height deviation from the mean line under the specified evaluation procedure |
| Rq | Root-mean-square profile-height deviation; it weights larger departures more strongly |
| Rz | A height parameter whose exact definition, evaluation and aggregation must follow the invoked standard edition |
| Other functional parameters | Peak, valley, skewness or material-ratio information selected for the surface's function |
One average cannot describe every functional feature. A surface can have an acceptable mean roughness and still contain an important isolated defect. Do not convert between Ra and Rz using a universal multiplier. ZEISS identifies Ra/Rz and additional functional parameters in its ISO 21920 guidance and emphasizes interpreting the actual surface specification. ZEISS, ISO 21920: Standard for Roughness Measurement, undated
ISO 21920-2:2021 addresses profile terms, definitions and parameters; ISO's listing identifies the June 2022 corrected version. Keep the full specification context rather than silently applying a new default to a legacy drawing. Instrument support for several standard modes makes mode selection part of the measurement record. ISO 21920-2:2021, official scope and edition information
Filtering, length and direction
Separate total traverse from the data length evaluated for the parameter. The stylus may need run-in/run-out travel as well as the evaluated region; a part that fits the nominal evaluation length may still lack usable access. Identify the profile, form removal, filter type, cutoff/nesting indices and evaluation length before comparing results.
ZEISS explains that measurement length, groove direction and filtering influence roughness results, and different nesting indices can produce different values. Use the specified direction and representative locations; where a method requires crossing the lay, record that orientation rather than relying on a visually convenient trace. Shortening a trace to fit a small land changes the measurement conditions and needs an agreed method. ZEISS, measurement conditions and filtering
Keep the raw profile and settings when investigating disagreements. A scalar value cannot show whether a scratch, contamination, edge crossing or levelling problem dominated the trace. Repeated measurements at the same location address different questions from measurements across representative locations.
Stylus geometry, force and access
The finite tip shapes the recorded profile. NIST's September 2, 2013 study of rectangular and trapezoidal roughness specimens explains that a tip can widen peaks and narrow valleys, changing Ra either upward or downward depending on profile shape. A larger tip does not simply produce a universally lower Ra. Inspect tip condition and choose the geometry required by the method. NIST, The Effect of Tip Size in Calibration of Surface Roughness Specimens with Rectangular Profiles
A bounded equipment example is the Mitutoyo SJ-220, 178-741-13, standard-drive 0.75 mN configuration. Catalogue E15036-US, printed February 2025, lists a 2 µm-radius, 60° tip, 17.5 mm traverse and 360 µm detector range. Its range/resolution entries include 360 µm/3.5 nm, 100 µm/0.9 nm and 25 µm/0.2 nm. These are range-dependent resolution figures, not uncertainty claims.
The same catalogue's 178-742-13 configuration uses 4 mN and a 5 µm-radius, 90° tip. The transverse-drive configurations have different traverse/access limits. Select by the exact drive and detector, and consider both stylus and skid loading on delicate surfaces; the catalogue lists skid force separately. Mitutoyo, Surftest SJ-220 catalogue
Calibration, uncertainty and use
Use clean, stable support and avoid vibration or relative motion during the trace. Preparation should remove contamination without changing the finish being inspected. NPL GPG37 covers positioning, measurement practice and uncertainty sources; a satisfactory reference-specimen check does not establish every possible surface measurement. NPL GPG37, Issue 2
Calibration must address the relevant measurement chain. NIST's stylus-profilometer description distinguishes traceability for height parameters from traceability for lateral spacing, with different references used for each. It also provides reference software/datasets for validating parameter calculations. A vertical check alone does not validate lateral scale or every software calculation. NIST, Stylus Profilometer
NPL GPG129, copyright 2013, discusses noise, flatness deviation, amplification, linearity and axis perpendicularity for areal contact stylus instruments. That scope illustrates why one “calibrated” label or finest display increment is insufficient; apply procedures appropriate to the actual instrument and profile/areal task. NPL, Calibration of the metrological characteristics of areal contact stylus instruments
Report parameter and units, standard/mode, trace location/direction, tip and force, reference arrangement, filter and evaluated length, repeated results and applicable checks. Include task uncertainty and the agreed decision rule when judging conformance near a limit.