Surface roughness Rt explained for metal parts and drawings

a[data-rs-seo-link]{text-decoration:underline!important;color:#1a56db!important;cursor:pointer!important;}a[data-rs-seo-link]{text-decoration:underline!important;color:#1a56db!important;cursor:pointer!important;}

What surface roughness Rt means

Surface roughness Rt is the total height of a roughness profile: the vertical distance from the highest measured peak to the deepest measured valley over the defined evaluation length. In practical terms, Rt answers a different question from Ra. Ra reports the average absolute roughness deviation. Rt reports the most extreme peak-to-valley span found in the measured trace.

That makes surface roughness Rt useful when one scratch, burr, pit, ridge, or machining mark could affect sealing, sliding, coating, fatigue life, or appearance. It is also why Rt should not be written as a bare number. The standard, filter, evaluation length, measurement direction, and acceptance rule all affect the reported value. Without those conditions, two inspectors can measure the same part and report different results.

For readers comparing multiple finishing terms, Breezcats keeps related articles in the surface finishes section. This article focuses on Rt as it appears in profile-based surface texture measurement for metal parts, machined components, formed parts, coated hardware, and general engineering drawings.

Rt in the standards context

Rt belongs to the family of profile roughness parameters. In profile measurement, a stylus or optical instrument records a line trace across the surface. Form and waviness are separated from roughness by the selected filtering method, and the remaining roughness profile is evaluated over a stated length. Standards such as ISO 21920-2:2021, ISO 21920-3:2021, and ASME B46.1 define the language, parameters, and specification approach used for this work.

As of November 2026, ISO 21920-2:2021 is the published ISO standard for terms, definitions, and profile surface texture parameters, with a second edition committee draft under development. ISO 21920-3:2021 covers specification operators for profile surface texture. This matters because many older drawings, instrument menus, and shop-floor practices still refer to ISO 4287 or older parameter names. When a drawing says only “Rt 6.3” with no standard or measurement conditions, the requirement is incomplete.

The basic idea of Rt is simple, but the measurement chain is not. A reported Rt value depends on the profile type, filter, cut-off or nesting index, evaluation length, stylus tip condition, measurement direction, and whether the surface was cleaned or deburred before measurement. For this reason, Rt should be treated as a controlled measurement result, not as a visual description of a rough or smooth finish.

How Rt differs from Ra and Rz

The most common mistake is assuming that Rt, Ra, and Rz are interchangeable. They are not. They may all be expressed in micrometres or microinches, but each parameter answers a different engineering question.

Parameter	What it mainly describes	Typical use	Main limitation
Ra	Average absolute profile deviation from the mean line	General process monitoring and broad finish comparison	Can hide isolated peaks, scratches, and valleys
Rz	Peak-to-valley height evaluated within sampling lengths or according to the stated standard	When height extremes matter more than an average, but some averaging is still useful	Definition and reporting can vary between ISO, ASME, and legacy usage
Rt	Total peak-to-valley height over the evaluation length	Detecting the largest height excursion in the measured trace	Highly sensitive to one defect, particle, burr, or scratch
Rmax or Rz1max	Largest single roughness depth in older or instrument-specific terminology	Legacy drawings and some inspection reports	Must not be assumed equal to Rt unless the specification says so

Ra is widely used because it is easy to compare and often stable for production control. However, an average can be forgiving. A surface may have an acceptable Ra value while still containing one deep groove that compromises a seal. Rt is stricter in this respect because it looks for the full vertical span of the measured roughness profile. If the highest peak and deepest valley are far apart, the Rt value rises even when most of the surface is relatively uniform.

Rz sits between these ideas in many practical discussions. It gives more attention to peak-to-valley height than Ra, but it is usually less dominated by one event than Rt when evaluated as an average across sampling lengths. Because the exact Rz definition has varied across standards and legacy conventions, drawings should not rely on shop understanding alone. The standard and revision should be stated.

Why Rt can change even when the surface looks the same

Two parts with similar visual finish can produce different Rt values. Rt is driven by the most extreme measured peak and valley in the selected trace. If the trace crosses a small nick, torn metal edge, burr, dragged abrasive particle, or deep machining groove, Rt can increase sharply. If a neighboring trace misses that feature, Rt may be lower.

This sensitivity is not a flaw in the parameter. It is the reason engineers use Rt. The value highlights extreme profile features that averages may reduce or hide. It becomes a problem only when the drawing does not define where and how to measure, or when inspectors compare results taken under different conditions.

Several measurement choices strongly influence Rt:

• Evaluation length. A longer length gives the measurement more opportunity to include an extreme peak or valley. A shorter length may miss a defect.
• Filtering. The selected filter separates roughness from waviness and form. Changing the cut-off can shift features into or out of the roughness profile.
• Measurement direction. Measuring across lay marks is often different from measuring parallel to them. Direction should match the functional risk.
• Surface preparation. Dirt, chips, oxide particles, coolant residue, or loose abrasive can create false high readings.
• Instrument condition. Stylus tip radius, tip wear, calibration, contact force, and vibration can affect the profile trace.

For high-risk surfaces, one Rt value from one trace is rarely enough to understand the part. A more robust plan may specify several traces, defined measurement zones, and complementary parameters such as Ra, Rz, Rp, Rv, Rsk, Rku, or material ratio parameters.

When Rt is worth specifying

Rt is most useful when the maximum peak-to-valley event has functional importance. It is less useful when the goal is only to monitor normal process variation or compare general surface smoothness. In hardware, machining, and metal finishing, Rt can be relevant in the following cases:

• Sealing surfaces. A deep valley or scratch can create a leakage path, while an excessive peak can damage a gasket or mating seal.
• Sliding or bearing contact. High asperities can accelerate running-in wear, scuffing, or localized contact stress.
• Coating and plating substrates. Extreme peaks may remain thinly covered, and deep valleys may trap process fluids or contamination.
• Fatigue-sensitive parts. Sharp valleys or scratches can become stress concentrators, especially on loaded metal surfaces.
• Visible decorative finishes. One pronounced line or pit may be unacceptable even if the average roughness is within a nominal range.

Rt is not always the right control parameter by itself. If the surface must retain lubricant, material ratio or bearing-area parameters may be more informative. If the concern is a directional machining pattern, spacing and lay information may be needed. If the surface is functionally three-dimensional, areal parameters such as Sa or Sz may be more appropriate than a single 2D profile trace. Rt should be selected because it matches a failure mode, not because it looks more stringent than Ra.

How to write an Rt requirement on a drawing

A clear Rt requirement should describe both the limit and the measurement method. The exact drawing syntax depends on the standard used by the organization, but the engineering intent should be complete enough for a supplier and an inspector to reproduce the result. See also: Buying Guides.

A practical checklist for an Rt specification includes:

• The parameter: Rt, written exactly as intended.
• The limit and rule: for example, maximum Rt not to exceed a stated value.
• The unit: usually µm or µin. Do not mix units in the same note.
• The standard and revision: for example, ISO 21920 series or ASME B46.1, as applicable.
• The profile and filter conditions: roughness profile, cut-off, nesting index, or other required filtering details.
• The evaluation length and sampling arrangement, where not already controlled by the standard or drawing.
• The measurement direction relative to machining lay or functional motion.
• The number and location of traces, especially on large, curved, or non-uniform surfaces.
• Any preparation rule, such as cleaning before measurement or excluding obvious handling damage only if the quality plan permits it.

An incomplete note such as “Rt 10 µm” may be understood inside one shop, but it is weak for supplier quality control. A better approach is to connect the value to a recognized standard, the intended surface zone, and the inspection conditions. This does not make the drawing longer for its own sake; it reduces disputes when parts are measured by different instruments or different organizations.

Common mistakes when using Rt

Treating Rt as a conversion from Ra. There is no universal, reliable conversion from Ra to Rt. Two surfaces can have the same Ra and very different peak-to-valley extremes. Any ratio between Ra and Rt is process-specific and should be validated with actual measurement data.

Assuming lower Rt is always better. A lower peak-to-valley height may be desirable for sealing or sliding, but some surfaces need texture for oil retention, coating adhesion, or controlled friction. The correct Rt limit depends on function.

Ignoring the standard revision. Older ISO 4287-based documents, ISO 21920-based documents, ASME B46.1 specifications, and instrument menus may not describe every related height parameter in the same way. The revision date matters.

Measuring in the easiest direction instead of the functional direction. A trace taken parallel to lay may produce a different result from one taken across lay. If the surface interacts with a seal, shaft, sliding part, or coating flow direction, measurement direction should be intentional.

Using Rt alone as an acceptance plan. Rt is excellent at finding extremes, but it does not explain their frequency, distribution, shape, or bearing behavior. For critical surfaces, pair Rt with other parameters and a visual or microscopic inspection criterion where appropriate.

Frequently asked questions

Is Rt the same as Rz?

No. Rt is the total peak-to-valley height over the evaluation length. Rz is a related height parameter, but its calculation and averaging depend on the standard and convention used. They should not be swapped on a drawing without engineering review.

Can surface roughness Rt be converted from Ra?

Not reliably. Ra is an average deviation, while Rt is controlled by the most extreme peak and valley in the measured profile. A real conversion would require validated process data for the same material, machine, tool condition, finishing method, and measurement setup.

Why does Rt vary more than Ra between measurements?

Rt varies more because it is sensitive to a single scratch, pit, burr, particle, or high asperity. If different traces cross different local features, the reported Rt values can change even on parts made by the same process.

Should a drawing specify Ra, Rt, or both?

Use Ra when the goal is general average roughness control. Use Rt when an extreme peak or valley can affect function. Use both when the process average and the worst local height matter. For critical interfaces, consider additional parameters instead of relying on a single number.

What should be checked before rejecting a part for high Rt?

Confirm the standard, instrument calibration, filter settings, evaluation length, measurement direction, surface cleanliness, and trace location. If the high Rt comes from handling damage, a burr, or embedded debris, the quality plan should define whether that feature is part of the measured surface condition or a separate defect.

The practical takeaway

Surface roughness Rt is valuable because it exposes the worst peak-to-valley height in a defined roughness trace. That makes it more sensitive than average roughness and more directly connected to certain functional risks. The same sensitivity also means Rt must be specified carefully. A useful Rt requirement names the parameter, limit, unit, standard, filtering conditions, evaluation length, direction, and inspection plan. When those details are clear, Rt becomes a practical control for surfaces where one extreme feature can matter more than the average finish.