Film thickness: both ends of the window fail
Film thickness is the most consequential variable in gelcoat work, and it is genuinely a window rather than a minimum. Too thin and too thick fail for opposite reasons.
Too thin fails through undercure. Air inhibition reaches a fixed depth into the film, so in a thin coat that inhibited zone is a large fraction of the total and the film never develops full hardness or crosslink density. An undercured film then alligators when the laminating resin's styrene attacks it. Thin films also let glass fibres from the skin coat approach the surface, producing fibre pattern and print-through, and they pinhole, which gives water a direct path into the laminate.
Too thick fails through brittleness and exotherm. Cured gelcoat is unreinforced, comparatively brittle resin. Excess thickness cracks and crazes, especially at corners and radii, and generates more exotherm, more cure shrinkage and more internal stress. Thick films also sag on vertical surfaces and are prone to pre-release from the mould.
The industry practice is to build to the supplier's specified wet film in several thin, even, overlapping passes rather than one heavy coat, and to measure with a notched wet-film gauge. A thickness that looks right by eye is very often too thin or too heavy.
The lamination window is a state, not a time
After the final pass the gelcoat must reach a specific condition before the skin coat goes on: firm and tack-free to a light touch, but not fully cured.
Laminate too early and styrene from the laminating resin attacks the still-soft film, producing alligatoring or wrinkling. Laminate too late and the air-inhibited back face has cured out, the chemical co-cure is lost, and you are left with a weak secondary bond that can show up later as delamination.
The shop test is a light brush or fingernail check: cured enough that a light brush leaves no trail, still tacky enough to mark with a fingernail. How long that takes depends on shop temperature, humidity, air movement, catalyst level and film thickness, which is exactly why it must be observed rather than timed from a previous job.
Mould preparation and release
The mould surface is the finish. Gloss is a replica of the tool, so any defect, wax build-up, dust or contamination transfers straight to the part.
Release agent quantity causes defects in both directions. Too much or unevenly applied release produces fisheyes through surface-tension imbalance, and encourages pre-release. Too little causes sticking and tear-out. Silicone contamination is the classic fisheye source and is notoriously hard to eliminate once it is loose in a shop.
Release systems are not freely interchangeable. Wax, semi-permanent chemical release and polyvinyl alcohol behave differently, and polyvinyl alcohol is well known for fisheyeing off a freshly waxed surface. Follow the release supplier's regime rather than shop habit.
What goes directly behind the gelcoat
The layer immediately behind the gelcoat is a structural and cosmetic buffer, not just more glass. A resin mist coat followed by a fine skin coat or surfacing veil produces a resin-rich, largely fibre-free layer that blocks the texture of the coarser reinforcement behind it and resists moisture penetration.
Standard practice is one or two layers of fine chopped strand mat behind the gelcoat before heavier fabric, specifically to stop print-through from the exotherm and shrinkage of the layers behind. Published guidance commonly calls for two layers of light mat before cloth and three before woven roving.
This is also where interface voids are created or avoided, which matters because a large share of service blisters originate at the gelcoat-to-laminate interface.
Temperature, humidity and the compressed air line
Temperature governs gel time, exotherm, cure completeness, viscosity and sprayability. Material temperature, mould temperature and shop temperature all count independently: a cold mould will undercure a film even in a warm shop. Excess heat shortens working time and drives porosity and cracking.
Humidity attacks through several routes. Water in the compressed air line is the single most commonly named remedy across troubleshooting guides, appearing against fisheyes, slow cure, poor gloss, chalking and water spotting. Condensation is the second route: solvent evaporation and spray-gun air cool the wet film, and if the surface drops below dew point, moisture condenses into it. Coatings practice is to stay at least three degrees above dew point. Water in the peroxide is a third, and is a named cause of wrinkling.
If a shop has an intermittent, hard-to-explain surface defect, draining the compressor and fitting a proper air dryer is the cheapest first experiment.
A defect table with general causes
Fisheyes, seen as circular voids down to the mould, come from surface-tension imbalance: silicone, oil or water contamination, excess or uneven release, viscosity too low, film too thin or a poor spray pattern.
Pinholes, tiny through-holes often revealed on sanding, are air or gas trapped at gel: poor atomisation, too thick a pass, high viscosity, gassy catalyst, excessive pump pressure or mould contamination.
Alligatoring and wrinkling come from styrene attacking an undercured film: laminated too early, film too thin, catalyst too low, shop too cold or cure uneven.
Print-through and fibre pattern come from the laminate's cure shrinkage and exotherm telegraphing to the surface: gelcoat too thin or undercured, fibres too close to the surface, no skin coat or veil, or demoulded too early.
Pre-release, where the gelcoat lifts before cure, comes from uneven or excessive thickness, over-fast cure, too much catalyst, delayed lamination, mould vibration or excessive release slip.
Cracking and crazing, worst at corners, come from excess thickness in high-stress areas, exotherm, cure shrinkage, demoulding undercured or insufficient laminate stiffness behind the surface.
Poor gloss is a mould problem first and a cure problem second. Chalking is ultraviolet light and moisture degrading the polyester and exposing pigment and filler, and it is accelerated by undercure. Sagging is thixotropy too low, film too thick or spray pressure too high. Poor hiding is film too thin or a pigment with insufficient opacity for that colour.
Catalyst level and post-cure
The correct peroxide level is a narrow band rather than a single number, and it moves with the peroxide brand and active-oxygen strength, the film thickness and the temperature. Both under-catalysing and over-catalysing are defect causes: too little gives undercure, a soft film and poor water resistance, while too much gives over-fast gel, pre-release, excess exotherm, cracking and porosity.
The practical discipline is to take the range from the data sheet and verify it with a gel-time test on the actual batch, at the actual shop temperature, rather than carrying a figure across from a different product or a different season.
Post-cure at controlled elevated temperature drives crosslinking further toward completion, raising hardness, heat-distortion temperature and chemical and water resistance while reducing residual styrene, which itself yellows under ultraviolet light. It is routine on tooling gelcoats and chemically exposed parts. The schedule must come from the data sheet, since an over-aggressive post-cure causes its own stress cracking.
Storage, settling and colour consistency
Pigment settling in the drum is normal. Mix to homogeneity before use and periodically during a long run, but do not mix continuously, because sustained shear destroys the thixotropic structure the product depends on. If a drum has been over-mixed, let it stand before use.
Gelcoat properties drift before a drum visibly gels: viscosity falls and cure time lengthens as the inhibitor package depletes. Store cool and out of direct sunlight, order to consumption rather than stockpiling, and re-check gel time on older stock.
On colour-critical work, use a single batch for a single structure where possible. Batch-to-batch variation is normally held to an instrumental colour tolerance, but processing adds its own shift on top of the supplier's, and a match under one light source can fail under another.