Quick answer
A cosmetic metal finish should not be specified by a color name and a phrase such as “smooth, uniform and free of defects.” A useful specification defines: the substrate and finishing route; approved master or range samples; controlled lighting and viewing conditions; color coordinates and a chosen color-difference equation; gloss geometry; texture and roughness requirements; cosmetic zones; defect type, size, count and spacing limits; and the sampling and lot-control plan. Instrument readings should support—not replace—visual approval of the assembled product.
No single ΔE, gloss or defect limit is correct for every product. Limits must be qualified on production-intent parts and agreed between buyer, finisher and manufacturer before mass production.
Why appearance specifications fail
Words such as black, satin, fine texture and no scratches sound clear until two people inspect the same part. A dark anodized housing can pass a color reading yet look different beside a neighboring extrusion. Two powder-coated panels can have similar average roughness but scatter light differently. A small mark that is unacceptable on a customer-facing bezel may be irrelevant on a hidden mounting flange.
Appearance is a system rather than a single number. The observed result depends on:
- alloy, temper, casting condition or sheet/extrusion source;
- machining, blasting, brushing, polishing and cleaning before finishing;
- anodizing, plating, painting, powder coating or physical-vapor-deposition conditions;
- coating thickness, cure and, for effect finishes, particle orientation;
- part geometry, orientation and adjacent materials;
- lighting spectrum, intensity and direction; and
- the observer, viewing distance and viewing time.
The specification must control the variables that affect the buying decision without creating limits that are impossible to measure reliably.
Start with the finished-part system
Identify the exact substrate and process stack before calling out appearance. “Black anodize” is not a complete finish specification. Record the aluminum alloy and temper, material form, pre-finish preparation, anodizing type, coloring route, sealing requirement, protected/masked areas and any performance standard that applies. For painted or powder-coated parts, identify pretreatment, coating family, color designation, cure requirement and performance class where relevant.
This matters because the finish can reveal rather than hide the substrate. Tool paths, weld heat tint, casting porosity, extrusion flow lines, polishing direction and different alloy lots can remain visible after finishing. The Aluminum Anodizers Council notes that anodized appearance is influenced by alloy, temper, gloss, surface structure and coloring process, and recommends agreeing an acceptable range for each project.
For an assembly with adjacent visible parts, also specify whether they must be made from the same material lot, finished in the same batch, or controlled as a matched set. A tolerance suitable for isolated parts may not prevent a noticeable mismatch when surfaces meet at a seam.
Define how the part will be viewed
Visual decisions are not reproducible unless the inspection scene is defined. Establish a viewing protocol that reflects normal customer use, then use the same protocol for sample approval, incoming inspection and supplier release.
At minimum, document:
- Illumination: lamp/source designation, light-booth model where used, approximate illuminance at the inspection plane, booth background and control of stray light.
- Part orientation: installed orientation and whether the evaluator may tilt or rotate the part.
- Viewing geometry: direction of illumination and observer relative to the surface.
- Distance and time: one value for each cosmetic zone if necessary.
- Observer requirements: normal color vision, training and correction for ordinary visual acuity.
- Cleanliness: a defined cleaning method and whether fingerprints, dust or protective-film residue are excluded before inspection.
ASTM D1729 provides a recognized practice for visual appraisal of diffusely illuminated opaque materials, but the product specification still has to name the conditions to be used. Daylight through a window, ceiling LEDs and a phone flashlight can produce different decisions, especially on metallic, brushed or low-gloss surfaces.
If the product will be sold under more than one important light source, evaluate the approved sample and production parts under each source. A pair that matches under one illuminant can separate under another, an effect commonly called metamerism.
Specify color as a measurement method, not only a target
The CIE L\a\b\ system describes lightness on L\, red–green position on a\, and yellow–blue position on b\. It gives supplier and customer a common language, but three target values are not enough for repeatable inspection.
A complete color callout should state:
- the approved physical standard and its identification number;
- target L\, a\ and b\* values, if numerical control is required;
- illuminant and standard observer;
- instrument geometry, aperture and measurement area;
- whether the specular component is included or excluded;
- the color-difference equation—such as ΔE\*ab or CIEDE2000—and all required settings;
- the permitted total difference and any separate directional limits for ΔL\, Δa\ or Δb\*;
- number and location of readings per part; and
- conditioning, cleaning, backing and orientation rules.
ISO/CIE 11664-4 defines the calculation of CIE 1976 L\a\b\* coordinates and Euclidean color differences. ASTM D2244 covers calculation of color tolerances from measured color coordinates and cautions that tolerances from different systems cannot be converted by one simple factor for all colors. Therefore, “ΔE ≤ X” is incomplete unless the equation and measurement configuration are stated.
Do not copy a familiar ΔE limit into a new program without qualification. Perceived acceptability also changes with color, gloss, texture, part size and adjacency. Build trial lots around the intended process, measure them, and have stakeholders rank the samples under the agreed viewing conditions. Use that evidence to set a capability-compatible limit.
Single-angle color data may also be insufficient for metallic-flake, pearlescent, interference or strongly directional finishes. These surfaces can change with illumination and viewing angle. Such projects may require multi-angle measurement plus physical range samples.
Control gloss at a stated geometry
Gloss measures reflected light near the specular direction; it is not a synonym for smoothness. Common coating methods use 20°, 60° or 85° geometries. ISO 2813 addresses gloss measurement at those geometries for non-textured coatings on plane, opaque substrates, while ASTM D523 covers specular-gloss measurement and references the same nominal geometries within its scope.
Choose the geometry during correlation trials and record it on the drawing or appearance standard. Do not permit one site to report 60° results while another uses 85°. Also define:
- target and tolerance in gloss units;
- instrument make/model or demonstrated equivalent;
- calibration and verification frequency;
- measurement direction for brushed or anisotropic surfaces;
- locations and number of readings; and
- treatment of curved areas smaller than the instrument footprint.
A gloss reading can miss haze, orange peel, waviness and reflected-image distortion. Those characteristics may need separate visual limit samples or dedicated appearance instrumentation. On bare or highly reflective metal, first confirm that the selected method and instrument correlate with the visual decision; a coating method should not be assumed valid for every metallic surface.
Separate roughness from visual texture
Surface roughness is useful process data, but Ra alone does not define what a customer sees. Ra averages profile-height deviations along a measured trace. Surfaces with different peak shapes, spacing, lay, waviness or isolated defects can share the same Ra and still look different.
If roughness is functionally or visually important, state the applicable drawing standard and the full measurement conditions: parameter, limit, cutoff/filter, evaluation length, stylus or optical method, trace direction and measurement locations. The ISO 21920 series provides current terminology and profile-based surface-texture parameters. NIST’s Surface Finish Metrology Tutorial likewise distinguishes height from spacing and explains the role of filtering and measurement method.
Then add the visual attributes that roughness does not capture:
- brushing or grain direction and continuity;
- texture scale and uniformity;
- orange peel, waviness or mottling;
- edge-to-field consistency;
- visible machining or polishing lines; and
- directional appearance after assembly.
For bead blasting, brushing and polishing, lock the media or abrasive specification, nominal process condition, direction, coverage and replacement controls where practical. Approve production-intent samples rather than a small flat coupon alone; geometry and operator access can change texture around corners, bosses and recesses.
Use master and boundary samples correctly
A “golden sample” is most useful when it is controlled like a measuring device. It should represent the intended appearance, not merely the best part ever produced. For attributes that naturally vary, an approved range—light/dark color, high/low gloss, or acceptable/unacceptable defect boundaries—is often more practical than a single master.
Create a register with sample ID, part revision, substrate lot, finish batch, approval signatures, date, storage location and retirement date. Protect samples from ultraviolet exposure, abrasion, fingerprints, chemicals and uncontrolled cleaning. Keep matched duplicates with the buyer and supplier, and compare them periodically. Photographs are useful for training and location marking, but monitors and camera processing make them poor sole color standards.
The written specification remains the authority. Samples clarify boundaries; they should not silently replace missing process, measurement or defect definitions.
Turn defects into zone-based acceptance criteria
Mark cosmetic zones directly on the drawing or an attached surface map. A practical structure is:
- Zone A: primary customer-facing or frequently handled surfaces;
- Zone B: visible during normal use but not a focal surface;
- Zone C: visible only during installation, service or deliberate repositioning; and
- Zone D: concealed after assembly.
The letters are not an industry-wide severity standard. Define them for the product and pair each zone with its own distance, viewing time and acceptance table. Published company guidelines, such as GE Vernova’s mechanical-component cosmetic inspection guide, illustrate the use of product-specific zones and viewing conditions; their numeric limits should not be copied as universal requirements.
Classify defects by mechanism and appearance: scratch, dent, pit, inclusion, blister, bare spot, stain, burn mark, rack/contact mark, weld print-through, streak, color nonuniformity or texture break. For each defect and zone, state allowable size, count, spacing and clustering rule. Also say whether limits apply per surface, per defined area or per entire part.
Do not write “defect-free.” Define exclusions for deliberately masked surfaces, contact points, cut edges or hidden tooling marks where appropriate. Safety, corrosion, coating adhesion and dimensional nonconformities should be controlled separately; a hidden bare spot can be functionally unacceptable even if its cosmetic zone is lenient.
Build measurement agreement and lot control
Before production approval, run a correlation study using representative parts near the acceptance boundary. Have trained appraisers at buyer and supplier sites inspect the same randomized set more than once. Compare within-appraiser repeatability, between-appraiser agreement and agreement with the approved reference decision. For instruments, compare readings on the same parts using the same settings, orientation and calibration practice.
Resolve disagreements before setting a narrow tolerance. Common causes include different apertures, included/excluded specular settings, curved-part positioning, dirty masters, lamp aging, measurement over mixed textures and inconsistent defect sizing.
The control plan should define a lot—for example by material lot, line, bath, paint/powder batch, shift and time window—and preserve traceability to process records. Specify first-off approval, sampling frequency, measurement locations, reaction rules, retention samples and escalation for trends. Track actual values, not pass/fail alone; drift toward a limit is visible earlier in L\*, gloss and process-control charts than in final rejection rates.
Practical specification checklist
Substrate and route
- What to record: Alloy/grade, temper, form, pretreatment, preparation, finish stack
- Qualification question: Does the approved sample use production-intent material and tooling?
Visual inspection
- What to record: Source/booth, illuminance, background, orientation, distance, time
- Qualification question: Do buyer and supplier reach the same decision?
Color
- What to record: Master ID, L\a\b\*, illuminant/observer, geometry, aperture, specular setting, ΔE equation
- Qualification question: Does the numerical limit predict visual acceptance?
Gloss
- What to record: Geometry, target/tolerance, locations, direction, instrument equivalence
- Qualification question: Is the surface within the selected method’s useful scope?
Texture
- What to record: Visual range, lay/grain, roughness parameter, filter, evaluation length, trace direction
- Qualification question: Can two surfaces with similar Ra still look different?
Defects
- What to record: Zone map, defect definitions, size/count/spacing, exceptions
- Qualification question: Are boundary samples available for subjective defects?
Sampling
- What to record: Lot definition, sample size/frequency, first-off and trend rules
- Qualification question: Will the plan detect drift before assemblies are built?
References
- What to record: Master/range IDs, revisions, storage and replacement
- Qualification question: Are identical controlled references available at both sites?
Limitations: what a cosmetic finish specification cannot do
A specification cannot remove normal process variation, make different substrates visually identical, or guarantee field appearance under every light source. Small coupons may not reproduce geometry-dependent coating flow, racking, polishing access or thermal history. Instrument values may not capture sparkle, grain, haze, waviness or the assembly-level contrast between adjacent parts. Accelerated color or corrosion testing also answers a different question from initial cosmetic acceptance.
Treat the first specification as a validated control method, not a permanent guess. Review it after pilot production and after any change to substrate supplier, alloy lot strategy, tooling, pretreatment, coating source, line, instrument or approved standard.
Send Alpha Metal the information needed for a finish review
For a practical feasibility review, send the part drawing and 3D file, substrate and finish route, annual volume, customer-facing zone map, assembly orientation, target color/gloss/texture, existing master or range samples, expected environment, functional coating requirements and current rejection examples.
Alpha Metal can help convert appearance intent into a production-ready part and inspection discussion. Contact chihwei@alphametalwork.com with your cosmetic metal component requirements.
Sources
- CIE, ISO/CIE 11664-4:2019 — Colorimetry—Part 4: CIE 1976 L\a\b\* colour space. https://www.iso.org/standard/74166.html
- ASTM International, ASTM D2244-25 — Standard Practice for Calculation of Color Tolerances and Color Differences from Instrumentally Measured Color Coordinates. https://www.astm.org/d2244-25.html
- ASTM International, ASTM E1164-23 — Standard Practice for Obtaining Spectrometric Data for Object-Color Evaluation. https://www.astm.org/e1164-23.html
- ASTM International, ASTM D1729-22 — Standard Practice for Visual Appraisal of Colors and Color Differences of Diffusely-Illuminated Opaque Materials. https://www.astm.org/d1729-22.html
- ISO, ISO 2813:2014 — Paints and varnishes—Determination of gloss value at 20°, 60° and 85°. https://www.iso.org/standard/56807.html
- ASTM International, ASTM D523-25 — Standard Test Method for Specular Gloss. https://store.astm.org/d0523-25.html
- ISO, ISO 21920-2:2021 — Geometrical product specifications (GPS)—Surface texture: Profile—Part 2: Terms, definitions and surface texture parameters. https://www.iso.org/standard/72226.html
- T. V. Vorburger and J. Raja, NIST, Surface Finish Metrology Tutorial, NISTIR 89-4088. https://www.nist.gov/publications/surface-finish-metrology-tutorial
- Aluminum Anodizers Council, Suggested Anodic Coating Specifications. https://www.anodizing.org/anodic-coating-specifications/
- GE Vernova, Cosmetic Inspection Guidelines for Mechanical Components, GA-SRC-0002 Rev. 2.0 (industry example of company-specific zone inspection; not a universal limit). https://www.gevernova.com/electrification/sites/default/files/2025-04/GA-SRC-0002%20Cosmetic%20Inspection%20Guidelines%20for%20Mechanical%20Components.pdf
- X-Rite, LAB Color Space and Values (plain-language explanation of L\, a\ and b\* axes). https://www.xrite.com/blog/lab-color-space
Image plan
Use the four third-party images only with the attribution and license handling in article2_images.json. Keep credits adjacent to each image or in a clearly linked image-credit section. Do not imply that photographed equipment is used by or endorsed by Alpha Metal. Crop only when permitted and preserve a record of modifications for ShareAlike compliance.
Discuss your enclosure or finish requirements with Alpha Metal
Share your drawings, target volume, substrate, finish requirements, and validation priorities. Contact Alpha Metal for a manufacturing review.
Sources, attribution, and intellectual property notice
- RAL CLASSIC K5 color fan. Image: Colourfeeling; CC BY-SA 3.0. Wikimedia Commons source and license details.
- Colorimeter for objective color measurement. Image: Sonal Shinde; CC BY-SA 4.0. Wikimedia Commons source and license details.
- Contact profilometer at LAAS-CNRS. Image: Guillaume Paumier; CC BY-SA 3.0. Wikimedia Commons source and license details.
- Fluidized-bed powder-coating schematic. Image: Dipl-ing-metaller; CC BY-SA 3.0. Wikimedia Commons source and license details.
Alpha Metal respects the intellectual property rights of creators and rights holders. If you believe that any material on this page infringes your copyright or other intellectual property rights, please contact us promptly and provide sufficient details to identify the material and your rights. We will review the request in good faith and, where appropriate, remove or amend the material without undue delay.
