How CD Texture Is Made: From the Disc Rainbow to Concentric Grooves in Metal

1. First, what CD texture actually is

Turn a CD over to the light and curved rainbows slide across the surface. Seeing something similar on metal — a phone key, a car knob, a decorative ring, the head of a decorative screw — the first reaction is usually "what colour is that plated". The answer is usually nothing: it is light that has been engineered.

CD texture means a field of equally spaced, fine concentric or spiral grooves machined into a metal or plastic surface. The name comes from its visual prototype, the fine tracks on a disc's data side. The trade also calls it concentric circle texture, sunburst texture or disc texture, usually meaning the same family — the naming describes what it looks like, not how it was made. One caution: sunburst sometimes refers specifically to the visual effect, rays radiating from the centre under a strong light, while the groove geometry remains concentric. The same microstructure can read as rings or as rays depending on the light source, so do not let the name steer the discussion.

It appears most often on rotationally symmetric or near-symmetric cosmetic faces: key tops, knob end faces, decorative rings, screw heads, camera bezels, watch crowns, door handle trim. Three functional benefits come along for the ride — grip, because the micro-grooves raise friction; fingerprint suppression, because diffuse reflection scatters the specular sheen of a print; and visual identity, because it produces a distinctive light-and-dark language at a given angle.

Distinguish it from the textures it is easily confused with, or the process discussion goes wrong later:

One line to separate them: hairline is lines, knurling is a grid, CD texture is rings, facet cutting is mirror facets. They all live on metal, and their visual logic has nothing in common.

2. Optics: a CD diffracts, metal reflects

This is the section most often assumed rather than understood, and it drives most of the engineering decisions.

A CD's rainbow is diffraction. The data track is a continuous spiral from the inner to the outer edge, with adjacent turns spaced at about 1.6 μm on a CD and about 0.74 μm on a DVD. That is the same order as the wavelength of visible light, roughly 0.38–0.78 μm, so the whole track behaves as a reflective diffraction grating. In directions satisfying the grating equation d · sinθ = mλ, one wavelength interferes constructively and brightens while others weaken, and different wavelengths leave at different angles. White light separates into a rainbow. The effect comes from periodic structure at wavelength scale, and has nothing to do with what data is stored.

A metal CD texture does not work that way. Engineered CD texture has pitch roughly in the 0.05–0.3 mm range, two orders of magnitude larger than visible wavelength — about 90 to 550 times the centre wavelength. At that scale the diffraction angle has collapsed onto the specular direction, so there is no rainbow. What you see instead is bands of light that move with viewing angle, and the mechanism is different: each groove flank is a small mirror, and its orientation rotates continuously around the circle. A point source therefore lights only the ring of grooves whose orientation satisfies the law of reflection, and you see a ring of light. A linear source such as a fluorescent tube stretches that ring into radial bands. Under diffuse or overcast light it degrades into a broad gradient.

This has a practical consequence: the premium look of CD texture is not colour, it is controlled anisotropic reflection. That is why it works without plating, and why it is acutely sensitive to two things — pitch consistency and flank roughness. Let one turn drift and the light ring shows a visible break; leave a scratch or a tool mark on a flank and the band immediately looks dirty.

One corollary: if you genuinely want a rainbow on metal rather than a grey-white band, the pitch has to come down to 1–2 μm, which is diffraction grating territory made by laser interference lithography or holographic mastering, not by routine turning. Products marketed as "rainbow CD texture" are usually coloured by a coating such as gradient PVD or an optical film, not by the texture splitting light.

3. Geometry: pitch, tooth form and depth are coupled

Many drawings say only "CD texture 0.1 mm". That is an incomplete specification. Once pitch P and tooth form are fixed, depth h is locked by geometry and the shop has no freedom left.

V profile. With included angle 2β, so half angle β, depth and pitch are related by:

The sharper the tooth, the deeper the same pitch cuts. The three common included angles:

R profile. A sharp tool chips quickly in production, so a radiused tool with corner radius R is more common. The surface becomes a series of scallops from overlapping passes, and depth approximates:

At P = 2R the scallop reaches the semicircular limit and depth equals R; beyond that a flat appears in the middle and the texture stops being continuous. Actual depths in μm for several corner radii:

The conclusion from both tables is direct: there are only two ways to make the texture deeper — increase the pitch or sharpen the tooth. Both make the process harder: a larger pitch coarsens the pattern and scatters the gloss, a sharper tooth weakens the tool, makes die crests chip and increases cracking after anodising. The mature approach is to fix depth and gloss first, then work back to pitch and tooth form, rather than writing a pitch and asking the shop to make it work.

One more caution: there is no national standard for CD texture. It is a cosmetic feature, and acceptance normally rests on the customer drawing plus a signed limit sample, with roughness assessed per the GB/T 1031 and GB/T 10610 framework. A drawing that says only "CD texture" is incomplete; it needs at least the pitch, the tooth form or corner radius, the diameter range covered, and a statement that the signed limit sample governs.

4. Four forming routes

There are not many ways to make CD texture, and four survive in production.

① Single-point diamond turning or fly cutting — one part at a time. The workpiece rotates and a single-crystal diamond tool advances along the face with a feed per revolution equal to the pitch, cutting concentric rings. Flat parts often use a large-diameter fly cutter instead, with the tool orbiting a large circle while the work feeds linearly; strictly, fly cutting leaves tangent arcs rather than true concentric rings, though a large enough cutter radius makes them nearly straight. Terminology is not consistent between shops, so confirm which path is actually used at the sample stage.

Diamond turning has clear capability limits. On diamond-machinable materials — aluminium, copper, brass, nickel-phosphorus — surface roughness typically reaches Ra ≤ 5 nm and at best Ra ≤ 2 nm, which is mirror finish straight off the tool with no polishing, and that is the physical reason the light band looks clean rather than dirty. Its limitation is just as clear: diamond cannot cut iron-based alloys directly. At cutting temperature the carbon in diamond has a strong chemical affinity for iron, causing catalytic graphitisation and diffusion wear, and tool life falls off a cliff. Steel and stainless CD texture therefore has three realistic routes: switch to CBN or coated carbide and accept lower gloss and shorter life; plate nickel-phosphorus first and then turn; or leave turning altogether and use a die. The limitation is material-level, not something a better ground edge can solve.

② Die coining or cold heading — make a die once, then copy. The texture is produced as a negative in the die cavity, and each part copies the die surface in a single forming operation. The process is properly called coining, a form of cold or closed-die forging: material is fully confined with almost no freedom to flow, pressure far exceeds yield, and the metal is squeezed into every fine detail of the cavity with minimal springback, reproducing sharp edges and fine patterns to micrometre accuracy.

Why it suits CD texture: the texture is a single feature on an end face, not something that has to be cut part by part. Once the die exists, the difference between part one and part one million is only die wear, and the unit cost approaches zero amortisation as volume rises. The precondition is a soft, ductile material — pure aluminium, 6063, brass, copper and low-carbon steel are the mainstays. Age-hardened conditions such as 6061-T6 must be annealed before forming and aged afterwards. Stainless has high flow stress and hardens fast, so die life and texture integrity need re-evaluating. The costs are real too: the texture in the die must itself be cut by turning, wire EDM or spark erosion, then polished to mirror and optionally hard chrome or DLC coated; the die crests are stress raisers, so CD texture dies live markedly shorter than ordinary forming dies; and the pattern is fixed — changing the texture means cutting a new die.

③ Rolling — the most productive route on surfaces of revolution. A textured roller presses into a rotating surface and the pattern is rolled on by plastic flow. Equipment is simple, cycle time is short and consistency comes from the roller. But concentric rings on an end face cannot be rolled, because the directional nature of rolling does not match radial symmetry, so this route mainly serves cylindrical and conical surfaces, with depth limited by how far the material will flow.

④ Etching or electroforming — the most pattern freedom. Photolithography plus chemical etching, or lithography plus electroforming to make a master and then replicate it. Any pattern is possible, with no rotational symmetry constraint. The costs differ: chemical etching suffers undercut, so sidewall verticality, depth uniformity and floor roughness are usually worse than turning or coining, and the extra masking step complicates yield and consistency; electroforming is a replication process whose accuracy depends on the master — the route used to make optical masters — but the process chain is longer and the cost higher.

Three further routes sit outside the main table but matter in specific cases: laser engraving, good for small batches, unusual shapes and personalised texture, but with a rough floor and a heat-affected zone that limit gloss; UV transfer or IMD, the plastics approach, which is still die replication and shares its logic with metal coining; and the most overlooked of all — making the texture in the die in the first place. The cavity texture is produced as a negative, usually by diamond turning a master and then electroforming or using it directly as the insert, or by wire EDM and spark erosion followed by polishing. Die texture accuracy sets finished texture accuracy, and the ceiling of that step is the ceiling of the whole line.

5. Materials: the same drawing is a different problem in each

The same drawing in a different material may need a completely different process route. Two variables dominate: hardness, which decides whether diamond turning and cold coining are possible, and ductility plus work-hardening tendency, which decides whether the texture reproduces faithfully.

Hardness values shift with condition — annealed, work hardened, aged — and different materials are quoted on their customary scales (HB, HV, HRC), so the numbers cannot be compared across rows; they indicate magnitude only. Confirm on a sample in the actual delivery condition.

In one sentence: aluminium is the best metal for CD texture — soft, thermally conductive, machinable to nanometre finish with diamond, and able to take colour and wear resistance from anodising. Stainless is the opposite — diamond cannot cut it, flow stress is high, work hardening is fast, and the same texture can cost several times what an aluminium part costs. That is why most stainless CD texture ends up either die coined or nickel plated and then turned.

6. Post-processing: half the texture is lost here

Cutting the texture is only half the job. The look depends on contrast between crest and trough, and almost every finishing process flattens some of that difference.

Anodising on aluminium. It is a conversion film rather than a coating: about 50 % of the film grows outward and about 50 % penetrates the substrate, so external dimensions grow by roughly half the film thickness. MIL-A-8625F states the dimensional increment as half the nominal film thickness, and the ISO 7599 and ISO 10074 systems likewise require film growth to be considered in tolerances, with the real split varying between 50/50 and 60/40 depending on alloy, bath and current density. Per GB/T 8013.1, anodic films are graded by average thickness: AA5, AA10, AA15, AA20 and AA25 at not less than 5, 10, 15, 20 and 25 μm average, with local minima of 4, 8, 12, 16 and 20 μm, and the current edition adds thinner AA3 and thicker AA30 and HA40 grades. Two systems dominate in practice: conventional sulphuric anodising at 5–25 μm, decorative parts usually 10–15 μm, microhardness about HV 200–350; and hard anodising at 25–100 μm, commonly 25–50 μm, microhardness about HV 400–600.

Three effects on CD texture, each capable of sending a batch back:

• Crests get rounded and contrast falls. The film covers the profile uniformly and blunts the sharp edges, and the thicker the film the greater the loss. So texture depth on aluminium must be set by what will remain after anodising — the usual practice is to cut one step deeper and correct against the first anodised sample.
• Dimensions move. Crests in the textured area grow by about half the film thickness. If the same part has a fit or a thread, the drawing must carry film compensation, or the problem only appears at assembly when rework is very expensive.
• Hard anodic film is hard and brittle. At sharp corners it is both difficult to cover completely, leaving bare metal, and prone to cracking from stress concentration. MIL-A-8625F table III requires at least R0.8 mm at sharp corners for 25 μm film and at least R1.6 mm for 50 μm. For a truly sharp V profile that is effectively a prohibition: a hard-anodised CD texture part needs radiused crests, or a different finish.

PVD and electroplating. PVD coatings — nitrides, carbides, DLC — are typically sub-micron to a few microns, cover uniformly, add hardness and brightness, and round the crests far less than a thick anodic film, which makes them the usual partner when both gloss and wear resistance are needed. Nickel and chrome plating also brighten, but throwing power in the troughs is worse than on the crests, so deep texture tends to come out bright on top and dark below; state this on the drawing in advance.

Oleophobic and anti-fingerprint coatings. Grooves collect fingerprints and oil, and once soiled the light band goes cloudy. Consumer electronics almost always add a nanometre-scale oleophobic layer, and its presence makes the comparison conditions for the limit sample important: acceptance has to be judged at the agreed light angle and viewing distance, or the same sample will be judged differently in two different shops.

7. Inspection and common defects

CD texture has a fairly standard set of inspection items, but every value depends strongly on the product and must be agreed in writing with the customer. The table gives the usual order of magnitude in the trade as a starting point for discussion, not as an acceptance standard.

The repeated "to be confirmed" is not a disclaimer, it is the fact of the matter: CD texture is a cosmetic feature and the judgement basis is almost always a signed limit sample plus agreed viewing conditions. Writing numbers on the drawing is only the first step. Yield is decided by whether both parties settle what counts as acceptable during sampling and then fix the viewing conditions.

One measurement trap: roughness on a CD texture face can only be assessed on a single facet, with a cut-off far shorter than the pitch. Measured with a conventional cut-off spanning several grooves, Ra is dominated by the groove geometry itself — tens of micrometres of depth will push the reading into the micrometre range, reporting depth rather than surface quality. A mirror-grade Ra requirement should therefore be accompanied on the drawing by the measurement direction and cut-off length.

Defect investigation works back from the symptom to the cause. The common set on the line:

8. How to choose: three questions

Everything above reduces to three questions.

• How many pieces a year? Up to a few thousand, turning wins — no die investment, and changing the texture means changing parameters. From tens of thousands, die coining total cost starts to beat turning. Above hundreds of thousands, amortisation is negligible and unit cost is just material plus forming.
• What material? Aluminium, copper, brass — all four routes are open, so choose on volume. Steel and stainless — turning is largely closed off by the diamond-iron reaction, so either plate nickel-phosphorus first or go to a die.
• How fine is the texture? Coarse texture at 0.2–0.3 mm pitch is well within coining and rolling. Fine texture at 0.05–0.1 mm makes cavity micromachining much harder and more expensive, so turning becomes the steadier option. Below 0.02–0.05 mm, essentially only turning can deliver reliably.

9. The economics: when a die pays for itself

The choice between turning and coining is a break-even calculation:

Worked through with assumed values. The numbers illustrate the method; substitute your own quotations and cycle times:

So below about twenty thousand pieces a year turning is usually cheaper and lets you change the texture without changing a die. At a hundred thousand a year, coining's unit cost advantage is already decisive. The crossover moves with part size, texture fineness and die life, but the order of magnitude holds: the economic turning point for CD texture sits around ten to twenty thousand pieces.

Two variables are easily forgotten. Die life: the fine crests of a CD texture die carry the whole contact stress, so it lives markedly shorter than an ordinary forming die, and once the texture dulls it must be refurbished or replaced — allow for that in the amortisation. Design change risk: turning changes the texture in a program, a die changes the texture by cutting a new die. If the product is still being restyled, cutting the die too early is often the most expensive decision available.

10. What the drawing has to say

Fill in every item and the drawing becomes executable, and quotations become comparable:

• ① Texture type and form — CD texture as concentric rings, or spiral, or straight; full face or an annular band.
• ② Pitch — value and tolerance, usually ±0.005–0.01 mm, and state that it is measured radially.
• ③ Tooth form — the included angle such as 90°, or the corner radius such as R0.1. These are alternatives; do not give two contradictory numbers.
• ④ Depth — check it against geometry: is it consistent with the pitch and tooth form given? Only an R profile allows depth to be specified independently.
• ⑤ Centre treatment — is a flat left at the centre, of what diameter, and is a machining remnant acceptable.
• ⑥ Outer edge — is a 45° bright chamfer wanted at the rim, and is a plain band left between chamfer and texture.
• ⑦ Material and condition — grade and delivery condition, for example 6061-O or 6061-T6. If a T6 part is to be coined, state that annealing before forming and re-ageing afterwards is permitted.
• ⑧ Post-processing — anodising grade and maximum film thickness, for example AA15 at 10–15 μm; whether PVD applies; whether an oleophobic coating is added. This is the key input to how deep the texture must be cut.
• ⑨ Fit compensation — if there are threads, bearing seats or mating holes, state who owns film compensation and at which operation it is applied.
• ⑩ Acceptance basis — write "per the limit sample signed by both parties" and agree the light source, angle and distance. Numbers without a signed sample guarantee an argument in production.

11. Summary: a forming problem, not a cosmetic one

CD texture is a textbook case. It looks like an appearance question and is really a precision forming question. The geometry locks the parameters, the material eliminates half the process routes, post-processing comes back and rewrites the texture, and acceptance depends on both parties stating clearly what they mean. Manage it as surface finishing and yield never stabilises. Manage it as a forming operation with geometric constraints and it stops being mysterious.

Which is why the three things worth settling early are: the geometric definition — pitch, tooth form, depth; the post-processing route, and how much texture it will consume; and the way volume filters the route — turn each part, or cut a die once. Settle those three and everything after is execution.

On the materials most often used for CD texture — aluminium alloy, brass and low-carbon steel — cold heading or cold extrusion in a single forming operation deserves early evaluation: the texture is defined once in the die and every part copies the same geometry, with pitch consistency guaranteed by the die rather than by tool condition. That is exactly the stability volume decorative parts need.

12. Sources and standards

• GB/T 8013.1, Anodic oxide coatings and organic polymer coatings on aluminium and its alloys, Part 1: Anodic oxide coatings: AA5, AA10, AA15, AA20 and AA25 average and local minimum thickness, sealing quality and corrosion test methods.
• ISO 7599, Anodizing of aluminium and its alloys — Method for specifying decorative and protective anodic oxidation coatings; ISO 10074, Specification for hard anodic oxidation coatings on aluminium and its alloys.
• MIL-A-8625F, Anodic Coatings for Aluminum and Aluminum Alloys, table III: edge and internal corner radii by nominal film thickness, 0.001 in (about 25 μm) to about 1/32 in (R0.8 mm) and 0.002 in (about 50 μm) to about 1/16 in (R1.6 mm); §6.10.1 gives the dimensional increment as half the film thickness.
• GB/T 1031, Geometrical product specifications (GPS) — Surface texture: Profile method — Surface roughness parameters and their values; GB/T 10610, GPS — Surface texture: Profile method — Rules and procedures for the assessment of surface texture.
• GB/T 6403.3, Knurling: types and dimensions.
• ECMA-130, Data Interchange on Read-only 120 mm Optical Data Disks (CD-ROM): track pitch 1.6 μm and its tolerance.
• Optics: the grating equation for reflective diffraction gratings, d · sinθ = mλ, with visible wavelength about 0.38–0.78 μm.
• Single Point Diamond Turning (SPDT) ultra-precision machining reference: typical non-ferrous surface roughness Ra ≤ 5 nm and best Ra ≤ 2 nm; the chemical wear mechanism that excludes diamond tools from iron-based materials; nickel-phosphorus plating as the alternative route.
• Cold and closed-die forging reference: coining: micrometre-level detail reproduction under triaxial compressive stress, minimal springback, ductility requirements and die wear characteristics.