It is photo-oxidation, not UV burning
Sunlight alone does comparatively little. The damaging process needs ultraviolet light and oxygen together, and it runs as a self-sustaining radical chain: initiation when a chromophoric impurity absorbs light, propagation when the resulting radical reacts with oxygen to form a peroxy radical, chain branching when the hydroperoxide formed splits into two further radicals, and eventual termination.
The chain-branching step is what makes weathering accelerate rather than proceed linearly. One absorbed photon can eventually produce many damaging radicals, which is why a sheet that looks fine for several seasons can deteriorate noticeably in the ones that follow.
It also explains why stabilisation works the way it does. An absorber reduces initiation by soaking up photons; a HALS interrupts propagation by scavenging radicals. Attacking one step alone leaves the other running.
Two weak points in a cured polyester
A cured unsaturated polyester is not one material but two joined together: a polyester backbone crosslinked through polystyrene bridges. Both are vulnerable, in different ways.
The ester carbonyl is the first. Excited carbonyl groups undergo direct cleavage of the backbone, and photo-oxidation of the carbon adjacent to the carbonyl converts it to an anhydride which then hydrolyses, cleaving the chain and leaving acid end groups behind. This is the route that ties sunlight exposure to both chain scission and surface bleaching.
The polystyrene crosslinks are the second, and they are the main source of yellowing. Photo-oxidation of those aromatic units forms quinone and related structures whose extended conjugation absorbs blue light. What the eye reads as yellowing, and later browning, is the accumulation of those oxidation products.
It is a surface phenomenon, and that changes everything
Studies of weathered glass-reinforced polyester find the products of photo-oxidative degradation confined to a thin subsurface layer, on the order of a few tens of microns. Everything a buyer complains about happens in that skin.
This has three practical consequences. It explains why a surface film or a resin-rich veil is disproportionately effective for its cost. It explains why an ultraviolet absorber, whose protection depends on light having already travelled through material, is structurally weak at exactly the depth that matters. And it explains why washing a chalked sheet restores much of its appearance: the damage really is only skin deep, until erosion exposes the layer beneath.
Why a weathered sheet can get stiffer while getting worse
This is the least intuitive part and it matters commercially, because stiffness is often used as informal evidence that a sheet is healthy.
Chain scission and additional crosslinking happen at the same time. Residual unreacted styrene continues to post-cure under service exposure, and trapped radicals recombine into further crosslinks. In laboratory work on glass-reinforced polyester, samples have gained measurable flexural strength during exposure while their surfaces were simultaneously eroding and exposing fibre.
So stiffness up does not mean healthy. Toughness and appearance move the other way, and impact and flexural retention, rather than hardness, are the properties worth specifying and tracking.
The damage sequence, symptom by symptom
Loss of gloss comes first and is the earliest measurable signal. Micro-roughening of the degraded surface scatters light that previously reflected specularly.
Yellowing follows, from the quinone and conjugated structures formed out of the aromatic units. It is measured as a colour difference against the original, not judged by eye.
Chalking appears as a powder that wipes off. Chain scission liberates low-molecular-weight fragments and leaves loose pigment and degraded polymer at the surface, and titanium dioxide's photocatalytic activity contributes directly to it.
Fibre bloom is the serious one. Resin erodes preferentially and glass does not, so the reinforcement gradually becomes prominent at the surface. Loss of light transmission, in translucent sheet, is the combined result of yellowing absorbing light, surface roughness scattering it and fibre bloom scattering it again at every exposed resin-to-glass interface.
Embrittlement and crazing complete the sequence, as net chain scission in the surface layer creates crack initiation sites.
Fibre bloom is a loop, not an endpoint
Once fibres are exposed at the surface, they create paths along which moisture wicks beneath the surface. Water at the fibre-to-matrix interface degrades the bond, which loosens more resin, which exposes more fibre.
That is a self-feeding cycle rather than a state that stabilises, and it is why the point at which fibre first becomes visible is a far more important milestone than it looks. It is also the strongest practical argument for keeping glass away from the weather face, through a resin-rich surface, a surfacing veil or a film, rather than relying on stabilisation alone.
How the industry measures all of this
Accelerated weathering uses either fluorescent ultraviolet cabinets or xenon-arc cabinets. Fluorescent testing, typically with UVA-340 or the more aggressive UVB-313 lamps combined with condensation cycles, is fast and good for ranking formulations. Xenon-arc testing reproduces the full spectrum including visible and infrared, and is the appropriate choice when colour is the property in question, because fluorescent lamps do not reproduce the visible spectrum well.
Natural exposure on outdoor racks remains the reference against which accelerated results are validated, and the two are not interchangeable.
The appearance properties are measured as instrumental colour difference, specular gloss at a stated geometry, and, for translucent sheet, haze and luminous transmittance. Cure state is checked by Barcol impressor hardness, and glass content by burn-off. Mechanical retention uses standard tensile and flexural methods.
Why accelerated hours do not convert into years
This is the question every buyer asks, and the honest answer is that no conversion exists. Acceleration factors cannot be calculated from first principles; they have to be measured, and a measured factor is specific to the two tests being compared, to the material, and to the particular property being tracked. A factor derived for gloss loss on one formulation does not transfer to colour change on another.
Accelerated testing does several things genuinely well. It ranks formulations against each other under identical conditions, detects gross formulation errors quickly, and supports quality control and batch consistency. What it does not do is predict calendar life.
Two further cautions are worth carrying into any supplier conversation. A weathering result is not even interpretable without its full operating conditions: which standard, which lamp or filter, what irradiance, what panel temperature, what wet cycle and what specimen construction. And accelerated methods explicitly exclude localised effects such as atmospheric pollution, biological growth and salt exposure, which in coastal or industrial sites can matter as much as the sunlight.