A two-pack kit is a stoichiometric calculation, not a packaging choice. Wrong ratio, wrong basis or an overrun pot life ruins the coat.
A two-component coating arrives as a kit, and the kit is a stoichiometric calculation rather than a packaging convenience. Part A carries the resin and part B carries the reactive groups that cross-link it. Change the proportion and the chemistry changes with it. Nothing downstream recovers the loss.
The ratio is chemistry, not viscosity control
Thinning a two-pack to improve flow is normal and permitted within the data sheet. Changing part A against part B is a different act entirely. Extra curing agent does not accelerate the cure in any useful way; it leaves unreacted amine or isocyanate looking for a partner that is not present, and the surplus shows up months later as blush, brittleness, or a chalky surface in service. Too little curing agent is worse: the film stays under-cured, retains solvent, and never reaches its specified adhesion or chemical resistance.
Ratios are stated either by volume or by weight, and the two are not interchangeable. Volume ratios assume the densities quoted in the sheet; weight ratios ignore them. Where a product is specified 4:1 by volume, weighing out 4:1 by mass delivers a different stoichiometry, sometimes by a wide margin. Read which basis applies before the first batch, not after.
Pot life begins at mixing, not at application
Pot life is the working window measured from the moment the two parts are combined, at the temperature stated. A batch mixed at eight in the morning and applied at one in the afternoon has spent five hours reacting in the can. The film it produces is not the film the data sheet describes, even though every step looked correct.
Temperature dominates that window. A product offering four hours at 20 °C may have barely one hour at 35 °C. In summer, on a hot deck or in direct sun, the practical window collapses faster than the crew expects. The answer is smaller batches mixed more often, not a longer clock.
Induction time is part of the instruction
Some epoxies require an induction period, commonly fifteen to thirty minutes, between mixing and application so the components can react and entrained air can escape. Skipping it is not cosmetic. The coating may solvent-pop, or develop a surface to which the next coat will not adhere, and the failure appears at the intercoat boundary long after the applicator has left.
How a dead batch announces itself
- Warmth in the can beyond ambient, with a rising smell of solvent or amine
- Visible thickening at the container wall before the bulk of the batch thickens
- Streaks or gel particles when the material is poured or rolled out
- Loss of flow that brushing cannot restore
Any of these means the batch has passed its window. There is no thinning it back to life. Adding solvent to a gelled epoxy produces a film with poor adhesion and unpredictable curing, and the defect will be traced to the coating rather than to the decision.
Controls that hold on site
- Measure by the basis stated in the data sheet, using a marked container or a ratio scale
- Record mix time and ambient temperature on the batch card, not on memory
- Divide the shift into batches sized to the pot life rather than to the area to be painted
- Keep both containers closed and clean; a lid left off part B draws moisture all day
- Discard rather than extend: the cost of one batch is trivial against a failed coat
None of this is difficult, and all of it is routine in a workshop with a written procedure. Failures cluster where the procedure lives in one person's head, usually the person who is on leave the week the surveyor arrives.