A coating that lifts off in one piece is the fastest way to lose a wood-finishing customer. It does not matter how good the gloss looked in the spray booth; the day a drawer front or a window sill gets scraped and the film comes away clean, the buyer stops blaming the substrate and starts blaming your formulation. We keep meeting this on customer lines, and in almost every case the peel was decided before the first coat ever dried. This walks through the seven root causes we actually see, and how to tell which one you have.
Before anything: confirm it is real adhesion failure
Half the "peeling" complaints we get are not adhesion at all. Score a cross-hatch lattice with a single-edge blade, lay ASTM D3359 tape over it, rip it off at 60° and read the release. If flakes come from the cut edges only, that is cohesive strength and internal stress, not bonding to the wood. If whole squares lift from bare substrate where there was no cut, you have a true interfacial adhesion problem. These two fail for different reasons, so run the test before you change a single ingredient.
Cause 1 — the substrate was never clean
Waterborne PU is far less forgiving of surface contamination than a solvent system, because it has almost no dissolving power to bite through oil. Sanding dust loaded with stearates, silicone release residue from the press plates, and finger oil all sit under the film as a release layer. Tannin-rich species like oak and teak add another problem: water lifts tannin and it bleeds into the coat. Wipe with a defoamed alcohol, abrade to a fresh surface, and if you run exotic species, seal first.
Cause 2 — nothing for the coating to grip
On dense hardwoods and on UV-cured or previously lacquered stock, there is no mechanical tooth. A waterborne film sitting on a glass-smooth factory finish will peel in one sheet. This is where a low-viscosity penetrating primer earns its keep. Our high-penetration primer resin YT-6035 runs at 50–200 mPa·s and 35 ± 2% solids, thin enough to wick into the open grain and lock a hard anchor film inside the substrate rather than on top of it.
Cause 3 — primer and topcoat were mismatched
Intercoat adhesion fails when the topcoat is a harder, more highly crosslinked film than the layer under it, or when the two use opposite stabilisation chemistries and refuse to coalesce across the boundary. Rule of thumb we follow: the topcoat should not be dramatically harder than the primer. Pair a penetrating anionic primer with a compatible anionic topcoat such as our high-hardness topcoat resin YT-6058, which reaches 1H–2H pencil hardness at 40 ± 2% solids without needing a hot solvent to knit down.
Cause 4 — the film never coalesced
This is the one people miss. Waterborne PU particles only fuse into a continuous film if they get enough time and temperature to deform. Spray a cold, humid line at 15°C with a fast flash and the particles dry as separate beads, stacked on top of each other. It looks like a film; under tape it pulls apart because there was never a continuous matrix to adhere. Extend flash-off, raise the dryer set point, or pick a grade with better low-temperature coalescence. You cannot crosslink your way out of a film that never closed. We saw it twice last winter, both on flooring lines running cold factories in January: one plant shortened the flash zone to chase a 15% line-speed gain, the other simply stopped pre-heating the boards. Panels looked fine at the stack and failed tape at the client’s QC lab weeks later. In both cases the fix was air and heat, not a new resin.
Cause 5 — too much hydrophilic group
Adhesion and water resistance pull in opposite directions in a PUD. The carboxylate groups that keep the dispersion stable and help it wet the surface also stay in the dried film and love water. Load them too high for a soft, tacky feel and any moisture that gets behind the coat re-plasticises the interface and releases it. If you chase water resistance, you usually give up some initial wetting; the fix is not more hydrophile, it is a crosslinker.
Cause 6 — no crosslinker, or the wrong one
A single-component PUD leans entirely on the resin for bond strength. A two-pack build locks the film to itself and to the substrate chemically. For wood we reach for an aliphatic non-ionic curing agent like waterborne curing agent YT-1102: 19–21% NCO, colourless, and a 5–6 hour open time that survives a full production shift. It raises hardness, water and chemical resistance and adhesion together, without the yellowing an aromatic hardener would cause on a light stain. Mind the pot life; a crosslinked film that gels in the line is a worse problem than the one you started with.
Cause 7 — film too thick, stress too high
Build a waterborne topcoat too heavy in one pass and shrinkage on dry-down stores stress across the whole sheet. The film wants to curl; the only place it can release is the interface, so it pops. Two thinner coats with proper flash between them beat one fat coat almost every time, especially on dimensional stock that swells and moves with humidity.
Which grade for which failure
Not every peel needs all three products. Match the fix to the diagnosis:
| Symptom | Likely cause | Start with |
|---|---|---|
| Lifts off in one clean sheet, no cut needed | No mechanical grip (Cause 2) | Penetrating primer YT-6035 |
| Topcoat separates from primer at the interface | Mismatched layers (Cause 3) | Anionic primer + YT-6058 topcoat |
| Film is chalky, breaks at the tape cuts | Never coalesced (Cause 4) | Fix flash-off / dryer before resin |
| Releases only after water or heat exposure | Too hydrophilic, no crosslink (Causes 5–6) | Add YT-1102 curing agent |
| Pops on edges and corners of thick coats | Internal stress (Cause 7) | Two thinner passes, more flash |
Prove the fix in the lab before you touch the line
We do not ship a "maybe". Run the candidate build through this before it goes near a customer batch:
- Cross-hatch + ASTM D3359 tape on your worst real substrate, not a perfect lab panel.
- Boil test: 2 hours in water, then tape again — this catches Cause 5 that a dry tape test hides.
- Cool-cabinet cycle for the low-temp coalescence question (Cause 4).
- Only after those pass, do a cost and line-speed comparison.
If you want the full range side by side, the waterborne PU resin lineup lists every grade with its TDS, and the wood coating application guide maps grades to primer, sealer and topcoat roles. For the VOC angle on the same substrate set, our piece on high-solids PUD wood coatings covers why we usually drop the heavy coalescent entirely.
Tell us the species, the line speed and how the coat is failing, and ask our application lab for a starting primer/topcoat/hardener combination. A crosscut photo of the failure is worth more than a hundred spec sheets.
