A freshly coated exhaust manifold peels off in sheets after the first heat cycle. The coating was expensive and the application looked careful, yet failure showed up almost immediately. The coating itself is rarely the problem—peeling almost always traces back to one of ten preventable mistakes in surface prep, compatibility, or application technique.
Surface Preparation and Corrosion Under the Film
Most peeling starts before the coating ever goes on. Rust, mill scale, old paint, or oil left on the substrate gives the coating something to bond to that isn’t the metal itself—and even a contamination layer too thin to see creates a weak interface that gives way the first time the part heats up and the coating and substrate expand at different rates.
The fix is grit-blasting or sanding to bare metal, ideally to a defined standard such as SSPC-SP 6 commercial blast cleaning, followed by a solvent wipe and full dry-down. Severely rusted parts need a chemical conversion coating first. Skipping this step is also how corrosion gets sealed in rather than kept out—a failure mode we cover in more depth in why high-temperature coating rusts too soon, since salt or residual corrosion left under a fresh coat drives the same lifting pressure from beneath. As a rule of thumb, budget roughly half of total application time for preparation, not application itself.
Substrate, Primer, and Coating Type Mismatches
Coatings don’t bond equally well to every metal. Bare aluminum, stainless steel, and cast iron all carry oxide layers that resist adhesion, so a coating can cure completely and still be only mechanically—not chemically—attached to the part. That distinction matters because mechanical-only bonds fail under thermal stress even when everything looked fine at installation. Aluminum generally needs a chromate conversion coating or epoxy primer underneath; stainless needs a primer formulated specifically for it; cast iron needs a wire-brush-and-etch step before anything goes on.
Primer compatibility compounds this: a primer not specified by the coating manufacturer can adhere well to the substrate while rejecting the topcoat, so the two layers separate from each other rather than from the metal. And coating type has to match the duty cycle—ceramic coatings hold up well under static high heat (furnaces, boilers, stacks) but crack under thermal cycling, while flexible silicone systems are built for the expansion and contraction of exhaust manifolds and headers. Our comparison of ceramic versus silicone high-temperature coatings and our guide to choosing coating by substrate both cover this selection process in more detail.
Thermal and Environmental Stress During Application
Steel expands at roughly 12 ppm/°C while many high-temperature coatings expand at 20–60 ppm/°C; across a 500°C temperature swing, that mismatch produces 0.5–2.4% strain that the coating has to absorb by stretching rather than cracking. Rigid, thick coatings can’t do that—thinner coats (2–4 mils rather than 6+) and inherently flexible chemistries hold up far better under repeated cycling.
Moisture is the other major environmental variable: coating applied over a damp surface, in high humidity, or onto a substrate still warm from a prior heat cycle traps water or drives off solvent too fast to form a sound film. Practical limits worth following—relative humidity under 85% during application, substrate temperature kept within roughly 90°F of ambient, and at least 30 minutes of dry-down after cleaning—prevent both failure modes at once.
Email Us if you’re troubleshooting a peeling failure and aren’t sure yet whether the cause is prep, compatibility, or application conditions—sending a few details usually narrows it down quickly.
Thickness and Cure Errors
Coating thicker than about 8 mils cures unevenly: the surface hardens while the interior stays soft, and that weak core delaminates once thermal cycling begins. Multiple thin passes (2–4 mils each), sanded lightly between coats, consistently outperform a single heavy application—more coating is not a substitute for correct coating.
Cure time causes a related but distinct failure. A part put into service before the coating has fully cured never develops its intended mechanical properties; heat just drives off the remaining solvent and leaves a weak residue behind, sometimes alongside the visible discoloration covered in why high-temperature paint changes color when heated. Room-temperature cure is slower than a heat-accelerated cure but generally produces the stronger final film—follow the data sheet’s cure window rather than shortening it under schedule pressure.
A Real-World Failure
An exhaust manifold gets coated with a generic “high-temperature paint”—actually an epoxy rated to 300°F, installed on a manifold that runs closer to 1,000°F. The coating peels within the first hour of engine operation. The root cause is a coating-selection mismatch, not a workmanship error: no amount of careful application saves a product rated for a third of the service temperature it’s exposed to. The fix is a full strip to bare metal, a ceramic coating rated to 2,000°F, and a full 48-hour cure before the engine runs again—after which the coating typically holds for five or more years of normal use.
If Peeling Has Already Started
Wire-brush away all loose material first, then inspect the exposed substrate for corrosion or contamination before doing anything else. Clean to bare metal again if needed, apply a conversion coating to prevent re-corrosion, and reapply per the manufacturer’s instructions. Adhesion can be verified before the part returns to service using ASTM D3359, the standard tape-test method for rating coating adhesion—cheap insurance against repeating the same failure. Give the repair a full, unaccelerated cure rather than rushing it back into service.
Peeling is almost always preventable, and every cause above traces back to preparation, compatibility, or cure discipline rather than the coating itself being defective.
Incure technical specialists can help diagnose a peeling failure and recommend the corrective coating system and application procedure for your specific substrate and operating conditions.
Contact Our Team to discuss a peeling coating failure and get guidance on prevention and repair.
Visit www.incurelab.com for more information.