Surface roughness is measured by dragging a fine-tipped probe across a surface or by using light-based instruments that scan without touching it. Both approaches capture the tiny peaks and valleys on a surface, then express the results as standardized numbers, most commonly Ra, which represents the average height of those irregularities in micrometers (µm) or microinches (µin). The method you choose depends on the material, the precision you need, and whether you can bring the part to a lab.
Key Roughness Parameters
Before measuring anything, it helps to know what number you’re actually after. Several standardized parameters describe roughness, and each tells you something slightly different about the surface.
Ra (Roughness Average) is the most widely used. It’s the arithmetic average of all the peak and valley heights along the measurement length. Ra is easy to compare across parts and processes, which is why it appears on most engineering drawings. A machined steel surface might have an Ra of 1.6 µm, while a polished one could be 0.1 µm or less.
Rz (Average Maximum Height) captures something Ra misses: extreme features. Rz divides the measurement length into sections, finds the tallest peak-to-deepest-valley distance in each section, and averages those values. Two surfaces can share the same Ra but have very different Rz values if one has occasional deep scratches.
Rq (RMS Roughness) is the root mean square of profile heights. Because it squares each value before averaging, Rq is more sensitive to outliers than Ra, making it useful when occasional tall peaks or deep valleys matter, such as in optical or sealing surfaces. Rq is typically about 11% higher than Ra for most machined surfaces.
Rt (Maximum Profile Height) is simply the vertical distance between the single highest peak and the single lowest valley across the entire measurement. It’s the worst-case number and is useful for applications where even one extreme feature could cause a problem, like a sealing gasket.
Contact Measurement: Stylus Profilometers
The most common way to measure roughness is with a stylus profilometer. A diamond-tipped probe, typically conical with a 2 µm tip radius and a 60° angle, is dragged across the surface under a light contact force (around 0.75 millinewtons). As the stylus rides over peaks and valleys, a feedback system tracks changes in the probe arm’s vertical position and reconstructs the surface profile.
Stylus profilometers come in two main forms. Portable (handheld) testers are compact, battery-powered, and designed for shop-floor or field use. You press them against a part, hit a button, and get an Ra reading in seconds. They’re ideal for in-process checks where you need a quick pass/fail answer. Benchtop profilometers sit in a lab, offer higher resolution, and can calculate a wider range of parameters. If your work involves R&D, calibration, or detailed surface profiling, a benchtop system is the better choice.
The trade-off with any stylus instrument is physical contact. The probe can scratch soft or delicate surfaces, and the tip itself wears over time, which affects accuracy. The tip radius also limits lateral resolution: features smaller than the tip simply can’t be detected. For hard metals and standard manufacturing tolerances, though, stylus profilometry remains the industry workhorse.
Getting a Good Stylus Reading
Place the instrument so the stylus travels perpendicular to the lay of the surface (the direction of the machining marks). Measuring parallel to the lay will ride along the grooves and understate the roughness. Make sure the part is clean and stable. Even small vibrations or debris will distort the profile. If the stylus can’t physically reach the measurement area, such as inside a deep bore, you can press a soft plastic film against the surface to create a replica, then measure the imprint instead.
Non-Contact Measurement: Optical Methods
When the surface is too soft, too small, or too valuable to touch, optical instruments measure roughness using light instead of a physical probe.
Confocal microscopy focuses a laser or white light beam onto the surface through a pinhole that blocks out-of-focus light. A version called chromatic confocal uses white light so that different wavelengths focus at different distances. By analyzing which wavelength returns in focus, the instrument determines the surface height at each point without moving the probe up and down. Commercial confocal systems achieve resolution down to about 12 nanometers, making them suitable for polished optics, semiconductor wafers, and microelectromechanical devices.
White light interferometry splits a beam of broadband light, bouncing half off the surface and half off a reference mirror. When the two beams recombine, they create interference patterns that encode nanometer-scale height differences. This method excels at measuring very smooth surfaces and can capture an entire area in a single scan rather than a single line.
Laser triangulation projects a laser spot onto the surface and uses a sensor at an angle to detect how the spot shifts as height changes. It’s faster and less precise than interferometry, making it a good fit for in-line quality checks on production parts.
Optical instruments are more expensive than stylus profilometers, but they avoid surface damage, measure faster over large areas, and can handle materials like polymers, coatings, and biological tissues that a diamond tip would deform or destroy.
2D Profiles vs. 3D Areal Measurement
Traditional stylus measurements produce a 2D profile: a single line traced across the surface. That line gives you Ra, Rz, and the other “R” parameters, but it only represents one narrow slice of the surface. A peak on a 2D profile isn’t necessarily a true summit on the actual surface, especially on randomly textured materials.
Areal (3D) measurement scans an entire patch of the surface and produces “S” parameters, the three-dimensional equivalents defined under the ISO 25178 standard. Sa is the areal version of Ra, Sq matches Rq, and Sz corresponds to Rt. But 3D measurement also unlocks parameters that don’t exist in 2D. Spatial parameters describe the directionality and pattern of the texture using autocorrelation functions. Hybrid parameters combine height and spacing information into a single number, such as Sdr, which tells you how much extra surface area the texture creates compared to a perfectly flat plane. This matters in applications like adhesive bonding or coating adhesion, where actual surface area drives performance.
Areal measurement is more reliable and more informative, but it requires optical or scanning-probe instruments and more processing time. For routine manufacturing checks, a 2D profile is often sufficient. For surfaces where function depends on texture in all directions, 3D measurement is worth the investment.
Units and Common Ranges
Roughness values are reported in micrometers (µm) internationally and in microinches (µin) in the United States. The conversion is straightforward: 1 µm equals roughly 39.37 µin, and 1 µin equals 0.0254 µm.
- Rough machining (milling, turning): Ra 3.2 to 12.5 µm (125 to 500 µin)
- Fine machining (grinding): Ra 0.4 to 1.6 µm (16 to 63 µin)
- Lapping and polishing: Ra 0.05 to 0.4 µm (2 to 16 µin)
- Superfinishing and optical polishing: Ra below 0.05 µm (below 2 µin)
Knowing where your process falls in these ranges helps you choose an instrument with the right resolution. A handheld tester that resolves to 0.01 µm is fine for ground surfaces but won’t tell you much about a polished optical lens.
Applicable Standards
Surface roughness measurement follows ISO standards that define parameters, measurement procedures, and filtering rules. The current standard for profile-based measurement is ISO 21920, which replaced the older ISO 4287 and ISO 4288 series. Part 2 of ISO 21920 covers terms, definitions, and surface texture parameters. For 3D areal measurement, ISO 25178 defines the “S” parameters and specifies instrument requirements. If a drawing or specification references Ra or Rz without naming a standard, ISO 21920 is the assumed framework in most industries.
Choosing the Right Approach
For most metalworking and manufacturing applications, a portable stylus tester delivers the speed and accuracy you need on the shop floor. If you need tighter control, a benchtop stylus profilometer gives you access to more parameters and better repeatability in a controlled environment. Switch to optical methods when you’re working with soft polymers, coatings, thin films, or any surface that a diamond tip could damage. And if your application depends on texture across an entire area rather than along a single line, invest in 3D areal measurement to get the full picture of what your surface actually looks like.