Performance properties & trade-offs

Color, gloss, hiding, mechanical, durability, corrosion, functional and rheological targets — a high-dimensional, interfering, non-additive objective space.

If processing is the operator and microstructure the hidden state, then performance is the output we actually grade. But it is not one number — it is a requirements vector of dozens of axes: a hue to hit, a gloss to hold, a film that bends without cracking, a barrier that keeps salt water off steel, a liquid that sprays today and survives the can for a year. The hard truth of this chapter is that these axes interfere: the knob that lifts one almost always drags another down. Before any mechanism (deferred to Physics) or any instrument (deferred to Measurement), we must first name and measure the axes — you cannot optimize a codomain you cannot coordinatize.

Colour and appearance: the optical axes

Appearance is the most visible block of the requirements vector, and the one where an engineer’s signal-processing intuition pays off fastest.

Colour is the perceived chromatic appearance, factored into three roughly independent axes: hue (which colour), chroma/saturation (how vivid), and lightness (how light or dark). A surface reflects a continuous spectrum — reflectance over roughly 380–730 nm, a high-dimensional analogue signal — which the eye samples with three cone types, a lossy 3-channel sensor.

The working coordinate system is CIE L*a*b*: L* lightness (\(0\to100\)), a* green(\(-\))/red(\(+\)), b* blue(\(-\))/yellow(\(+\)), engineered to be approximately perceptually uniform. The distance ΔE (the Euclidean ΔE76, or the better-shaped ΔE2000) is a perceptual loss function — \(\Delta E\approx 1\) is roughly a just-noticeable difference — so colour matching is loss minimization toward a target \((L^*,a^*,b^*)\). The catch: the “weights” you turn are pigment loadings, and they move every other property below (a coupling we return to as P8).

Gloss and haze describe how light leaves the surface. Gloss is the tendency to reflect specularly — mirror-like, in a narrow lobe; haze is milky scattering near but not at the specular direction. This is a surface-roughness (spatial-frequency) effect: a smooth film returns a coherent specular spike (high gloss), roughness broadens the lobe (matte), and gloss is essentially the ratio of specular to diffuse return. It is reported in gloss units (GU) at a fixed geometry — 60° for mid-gloss, 20° for high-gloss, 85° for matte — with DOI (distinctness-of-image) measuring how undistorted a reflected edge looks, the figure of merit for automotive clearcoats.

The remaining optical axes round out the picture:

• Opacity / hiding power and tint strength. Hiding is signal attenuation — suppressing the substrate so it can’t be read through the film — quantified by contrast ratio (reflectance over black versus white substrate; \(\approx 1.0\) is fully hiding). Tint strength is a gain: colour shift per unit colorant. White hiding leans on high-refractive-index \(\ce{TiO2}\) scattering; coloured and black hiding on absorption. More pigment hides better but thickens the film, raises viscosity, and can embrittle it — and pushing pigment too far crosses the CPVC and collapses gloss, adhesion, and toughness at once.
• Transparency and translucency are the controlled non-hiding case: a low-loss, low-blur channel that lets the layer below show through, optionally tinted like a coloured filter. Clearcoats are the canonical transparent layer; candy colours use transparent colorants that absorb without scattering. The enemies are haze (unwanted scatter) and yellowing, fought by dissolving or nano-refining components and refractive-index-matching binder to additives.
• Metallic, pearlescent, and goniochromatic (“flip”) effects make reflected colour or brightness depend on viewing angle. A normal pigment is angle-independent (scalar reflectance); these are angle-dependent, BRDF-rich materials whose output is a function of geometry, like a phased array that steers with angle. Metallics use aluminium flakes (tiny mirrors, giving lightness “flop”); pearlescents use mica/oxide platelets (thin-film interference); goniochromatic flakes shift hue with angle. They need multi-angle colour (L*a*b* at, e.g., 15°/45°/110°), and — crucially — appearance depends on flake orientation, set by application and drying flow: a process-coupled variable.

The automotive stack: decomposing the vector across layers

You don’t ask one layer to satisfy the whole requirements vector. A modern car finish is a designed stack of functional layers, each owning one responsibility with a clean interface — adhesion — to its neighbours.

As the figure shows, the stack runs bottom-to-top from bare metal to the visible surface, each layer solving a sub-problem the layer above or below should not have to.

• Substrate — the metal (or plastic) being protected.
• E-coat / anticorrosion — an electrodeposited primer; the primary corrosion defence and adhesion to bare metal.
• Primer / surfacer — smooths the surface, adds chip resistance, blocks UV from reaching layers below.
• Basecoat — carries the colour and effect pigments; optimized for appearance, not toughness.
• Clearcoat — the transparent top: gloss, depth, and UV / chemical / scratch resistance.

The basecoat + clearcoat together form the familiar two-layer appearance system. But layers must be mutually compatible — inter-coat adhesion, no lifting — and co- or sequentially-cured without harming each other.

Mechanical integrity: holding together under load

Once cured, the film is a thin structural member, and engineers can read its mechanical specs as familiar reliability ratings. The recurring antagonism here is hardness versus flexibility — a textbook Pareto pair, because both are governed by the same knob, crosslink density.

Two deserve a closer look. Adhesion is interfacial — the bulk film can be flawless, but if the bond to substrate or previous layer fails, the system fails. It is notoriously sensitive to contamination and often degrades after water or heat exposure (wet adhesion), so dry adhesion passing tells you little. Scratch resistance shows a clever escape from the hardness wall: premium clears chase elastic recovery — marks that relax back out — sidestepping the hardness-versus-flexibility trade instead of paying its full price.

Durability and protective function: surviving the duty cycle

The protective axes are long-term reliability under a harsh duty cycle — the coating’s MTBF under combined thermal, radiation, chemical, and electrochemical stress — plus a growing set of functional requirements driven by sensors and energy.

Weathering is aging under prolonged sun, moisture, heat, and oxygen. UV photons break polymer bonds over time, producing chalking (binder erodes, leaving loose pigment), gloss loss, colour fade (ΔE drift), cracking, and delamination. It is assessed by accelerated QUV / xenon-arc testing plus real-world exposure, with UV absorbers and HALS stabilizers extending life. The deep difficulty is extrapolation: predicting multi-year field behaviour from weeks of accelerated testing, across an imperfect accelerated-to-real transfer function — passing QUV is not the same as field durability, and lamp spectrum matters.

Chemical / stain / solvent resistance is robustness to hostile inputs — fault tolerance against adversarial environmental “data” (acids, fuels, cleaners, food). It is probed by spot/immersion exposure and solvent double-rubs (MEK rubs until the film softens), which double as a cure check: an under-cured film fails the rub even when everything else is right. Resistance rises with crosslink density, so it tracks hardness and works against flexibility — and “resist everything” is impossible.

Corrosion protection keeps the metal substrate from corroding, via three defensive strategies that map cleanly onto layered security:

Real systems combine all three (a zinc-rich primer is sacrificial + barrier, under a barrier topcoat), and the e-coat is the primary defence. Performance is screened by salt-spray (rust creep from a scribe over hundreds-to-thousands of hours) and electrochemical methods. Corrosion couples hard to wet adhesion and film integrity — a single pinhole or scratch localizes attack — and the mechanism choice dictates ingredients, including a sharp regulatory edge: inhibitive chromates are now restricted, forcing reformulation. (Salt-spray hours are a comparative screen, not a literal service-life number.)

The functional axes are newer and increasingly sensor-driven:

• Heat-shielding / IR-reflective (“cool”) coatings are spectral filtering on the energy budget. Roughly half of solar power sits in the near-IR, invisible to the eye, so a cool coating is a band-selective mirror: reflect NIR (reject heat) while still hitting the visible colour target — quantified by total solar reflectance (TSR). The trick is making even a dark coating reflect NIR, so a “black” roof runs far cooler than conventional black: two spectral constraints on one pigment system, often in tension.
• Radar / LiDAR / mm-wave transparency is EM compatibility for the autonomous-vehicle sensor stack. Paint over radar ($$76–81 GHz), LiDAR (NIR $$905/1550 nm), or cameras must be low-loss where it covers a sensor. The conflict is acute: the metallic flakes that give the prized “flip” are conductive and attenuate radar — directly contradicting the metallic-look axis.
• Antifouling / self-cleaning / hydrophobic surfaces are designed boundary conditions so unwanted inputs don’t stick, set by surface energy and texture. Superhydrophobic surfaces bead and shed water (the lotus effect); hydrophilic ones sheet water to wash dirt away. But the low-surface-energy additives that shed water also hurt recoatability and inter-coat adhesion.
• Conductivity / EMI shielding turns an insulator into a controlled sheet-resistance or shielding layer — familiar from EMI enclosures and ESD-safe materials. Conductive fillers (carbon black, metal flake or fibre) must exceed a percolation threshold so a connected network spans the film; specs are sheet resistance (Ω/sq) and shielding effectiveness (dB). Percolation is the same graph-connectivity transition that governs gelation and cure.

The process window: the uncured product must apply, then survive the can

Half the requirements vector concerns the liquid, before any of the cured properties exist. A coating that cannot be applied — or that has spoiled in storage — never reaches the substrate, so these are first-class constraints, not afterthoughts.

Storage / shelf stability is data-at-rest integrity: the metastable formulation must not drift or corrupt before use. Failure modes are settling (a hard cake), syneresis, viscosity drift, gelling, skinning, and microbial spoilage — and the very components that deliver performance (reactive crosslinkers, dense pigments) are the ones most prone to react or settle. It is probed by accelerated heat-aging and freeze-thaw cycling; a formula perfect fresh can be useless after six months.

Viscosity and the application window define the operating envelope — the input dynamic range outside which the process fails regardless of the formula. Coatings are typically non-Newtonian (shear-thinning): viscosity is a function of shear rate, not a single number. Shear-thinning is desirable — low viscosity under high shear (it atomizes or brushes), high viscosity at rest (it won’t sag or settle) — and the window spans temperature, humidity, thickness, and recoat timing. The real spec is the rheology curve.

Sag versus leveling is the classic process-window trade-off, and it reads as a damping problem: too little flow leaves permanent ripples (under-damped texture), too much runs and drips on a vertical surface (over-damped slump). Both are governed by the same low-shear viscosity over the same drying interval, so improving one worsens the other — they share a mechanism and must be co-optimized. The escape is time-dependent rheology (thixotropy): flow briefly, then build viscosity fast.

Pot life / open time is a countdown timer started at mixing. For reactive 2K systems (isocyanate + polyol; epoxy + amine), once the two parts meet, crosslinking climbs toward gelation — a hard deadline. Pot life is how long the mixed material stays usable (minutes to hours); open time, how long the applied film stays workable. Faster cure buys shorter pot life: a throughput-versus-working-time trade.

The multi-objective, interfering nature

Everything above converges on one structural fact: a coating must satisfy many properties at once, the knobs that lift one usually drag another down, and so there is no single best answer — only a set of non-dominated compromises.

As the figure shows — coating specs plotted as durability versus gloss/chroma — the achievable region has a frontier, and feasible designs pile up against it; you buy more durability only by spending gloss or chroma.

The interference is not bad luck — it is structural, because the objectives are coupled through a handful of shared variables: crosslink density, pigment volume fraction, low-shear viscosity, surface energy. Turning any one moves several objectives together:

Two more synthesis problems sit underneath the Pareto behaviour. First, non-additivity (P8): properties emerge from interactions of components — output \(\ne \sum\)(ingredient effects), with strong cross-terms, like a heavily coupled circuit where adding one component shifts the operating point of everything. A single co-solvent added for flow shifts drying, which re-orients flakes, which moves colour. This coupling is why the objectives interfere — shared variables and cross-effects are the mechanistic root of the front, and the reason reformulating for one new constraint ripples through rheology, drying, appearance, and stability rather than yielding to a local patch.

Second, sensory versus instrumental targets (P9): many real requirements arrive in human language (“luxurious,” “warm,” “deep ruby-like,” “high-grade texture”), while the lab speaks in spectra, ΔE, and gloss units.

Recap

• Performance is the codomain — a requirements vector, not a scalar. Per P13 we coordinatize it first (descriptors), deferring mechanism to Physics and instruments to Measurement.
• The axes group into themes: optical (colour, gloss/haze, hiding, transparency, flip), mechanical (hardness, flexibility, impact, adhesion, abrasion), durability/functional (weathering, chemical, corrosion, cool, radar, antifouling, conductivity), and the process window (storage, viscosity, sag/leveling, pot life, VOC).
• The stack decomposes the vector spatially — substrate / e-coat / primer / basecoat / clearcoat — so system properties (inter-coat adhesion) dominate layer properties (P8).
• The objectives interfere (P5) because they are coupled through shared variables (crosslink density, pigment volume, low-shear viscosity, surface energy); the feasible region has a Pareto front, and formulation is choosing a point on it to match an application’s weights.
• Properties are non-additive (P8): the film is a coupled nonlinear system, output \(\ne\sum\)(parts) — the mechanistic reason the front exists.
• Targets are dual (P9): perceptual briefs and instrumental readings of one physical state, reconciled by a learned, scope-dependent mapping; the meter is necessary, the eye is final.
• Bridge: the trade-off surface is an antichain in a product order — order theory, not containment algebra.