Surface Roughness Measurement: Methods and Reports

Two labs measure the exact same milled surface and land on noticeably different Ra values, even though both worked correctly. The reason is almost always a different measurement setting on the roughness tester, most often the cut-off wavelength.

In brief: Surface roughness measurement determines the microstructure of a part surface, most commonly through the tactile stylus method or through contactless optical systems. What actually decides whether a result is comparable is the chosen cut-off wavelength (0.08 to 8 mm) and the evaluation length. For buyers, what matters in the end isn't the raw Ra value itself but whether the report discloses the measurement conditions behind it.

How Does Tactile Roughness Measurement With the Stylus Method Work?

In the stylus method, a fine diamond tip moves across the part surface at a constant speed and traces its height profile. The evaluation software then calculates parameters like Ra or Rz (mean roughness depth) from that profile. These instruments are often called roughness testers or, colloquially, by the genericized brand name Perthometer, after one of the first commercial manufacturers of such devices. The process breaks down into three steps:

• Tracing: The diamond tip moves at a constant speed across the defined evaluation length.

• Filtering: The software uses the cut-off wavelength to separate roughness from waviness components.

• Evaluating: From the filtered profile, the software calculates the requested parameters, such as Ra or Rz.

Assessing surface texture through stylus measurement has long been firmly established in manufacturing quality assurance and remains the most common surface roughness measurement method, according to Germany's national metrology institute PTB in its report on the new ISO 21920 standard series.

For most parts, the technique counts as the reference against which optical instruments are also checked.

In our quality discussions with manufacturing partners in the Far East, we specifically ask, for critical surfaces, which stylus tip radius and which cut-off wavelength were used for the measurement. Without that detail, a supplied Ra value is hard to compare against your own drawing requirement, even when both numbers look identical at first glance.

What Role Do Cut-Off Wavelength and Evaluation Length Play in the Result?

The cut-off wavelength separates the fine roughness components from the coarser waviness components in the measurement signal, much like a sieve separating fine sand from coarse gravel. The rougher the expected surface, the wider the cut-off wavelength chosen.

The cut-off wavelength follows a fixed value series from 0.08 through 0.25, 0.8, and 2.5 up to 8 mm. The evaluation length is typically made up of five individual sampling lengths of that same width. A ground surface with an expected Rz around 7 µm, for example, gets evaluated with a cut-off wavelength of 0.8 mm and a resulting total evaluation length of 4 mm. If the cut-off wavelength is chosen too narrow for a coarse milled surface, waviness components get mistakenly counted as roughness. The result then gets pushed upward.

One detail has changed with the new ISO 21920 standard series. Parameters are now mostly calculated directly over the full evaluation length rather than first per individual sampling length. That produces small differences in some cases between old and new results measured on the same surface.

When Does Optical Measurement Pay Off Over Tactile Measurement?

Optical, contactless roughness measurement doesn't trace the surface mechanically. Instead, it captures it through methods like focus variation, white-light interferometry, or confocal microscopy, with no mechanical contact at all. That means no contact pressure that could damage sensitive or coated surfaces.

In sourcing, we often see buyers assume optical measurement systems are automatically the more accurate method, since they capture an entire area instead of just a single line. For a straightforward Ra or Rz check against a drawing, the tactile method remains the reference. Optical systems deliver their real advantage when the full topography of a surface matters, not just one single parameter.

For optical surface measuring instruments, which capture an entire area instead of a single line, a separate standard series, DIN EN ISO 25178, governs the terms and parameters for areal surface texture, especially relevant for complex, three-dimensionally structured surfaces.

What Does the 16% Rule Mean for Evaluating a Measurement?

The 16% rule is one of the possible tolerance acceptance rules a lab can use to decide whether a measured surface counts as compliant with the drawing overall. Under this rule, a surface passes if no more than 16 percent of the individual measured values exceed the permitted limit, rather than judging every single value in isolation.

What matters for buyers is that a drawing should explicitly state which rule applies. The specification operators for roughness measurement, including the evaluation rule, are governed by DIN EN ISO 21920-3. Without that specification on the drawing, the testing lab picks the evaluation rule at its own discretion. That can lead two suppliers working from the same drawing to reach differently strict release decisions. Similar to geometric tolerances, where reference points also need to be defined unambiguously, a clear drawing specification beats an implicit assumption here too.

What Should Be in the Roughness Measurement Report, and What Should Buyers Check?

A meaningful measurement report doesn't just document the measured parameter itself. It also documents the conditions under which that value was produced.

A report that states only the bare Ra value carries little weight for critical parts. Similar to ISO 2768 for general dimensional tolerances, it pays to fix the testing conditions already in the technical inquiry instead of leaving them to the manufacturing partner.

For safety-relevant parts, this requirement pairs well with an inspection certificate type 3.1, which confirms the measurement was taken by an independent inspector on the actually delivered material. A first surface check also belongs in regular incoming goods inspection, even though that step usually measures only on a sample basis, not with lab-grade precision.

Frequently Asked Questions About Surface Roughness Measurement

How does a surface roughness measurement work?

In tactile roughness measurement, a diamond tip travels across the workpiece at a constant speed and records its height profile. From that profile, the evaluation software then derives parameters like Ra or Rz for the defined evaluation length.

What's the difference between tactile and optical roughness measurement?

The contact method mechanically traces the workpiece with a diamond tip and serves as the reference for Ra and Rz values per drawing. Optical instruments work without touching the part and capture an entire area instead, gentler on delicate finishes, but not always directly comparable for a straightforward parameter check.

What does cut-off wavelength mean in roughness measurement?

The cut-off wavelength separates the fine roughness components from the coarser waviness components in the measurement signal. It's chosen from a fixed value series between 0.08 and 8 mm, depending on the expected roughness.

Which standard governs surface roughness measurement?

Profile-based roughness measurement is governed by the three-part ISO 21920 standard series, which merges the earlier standards ISO 1302, ISO 4287, ISO 4288, and ISO 13565-2/-3. For areal, optical measurement methods, the ISO 25178 standard series additionally applies.

What should a roughness measurement report include at minimum?

A solid report should specify, besides the parameter itself, the cut-off wavelength used, the evaluation length, the measurement method (tactile/optical), and the evaluation rule applied. Without these details, a result is hard to compare against any other test.

Conclusion: Surface Roughness Measurement Only Delivers Comparable Values With Clear Conditions

An Ra or Rz value alone says little if the cut-off wavelength, evaluation length, and evaluation rule are missing from the report. Buyers who fix these conditions already in the technical inquiry can actually compare measurement results across manufacturing partners, instead of relying on a single number.

Line Up supports buyers in specifying surface requirements precisely and in assessing the measurement reports manufacturing partners in the Far East deliver. 👉 Schedule a no-obligation consultation and find out which measurement conditions your next part actually needs.