A plastic-to-plastic epoxy joint that fails in testing or in the field leaves a fracture signature that points almost directly at the root cause — reading that signature correctly saves a round of guesswork before the next batch of parts goes out the door.
Signature: A Clean Release From One Plastic Surface, Adhesive Fully Intact
When cured epoxy peels cleanly off one of the two plastic faces, leaving a smooth, unmarked surface behind, the bond never chemically anchored to that substrate in the first place. This is the signature of an untreated low-surface-energy plastic — polypropylene, polyethylene, or PTFE — that needed flame, corona, or plasma activation before bonding and didn’t get it, or got it too long before bonding and lost the effect. Surface energy on activated polyolefins decays over hours to days depending on storage conditions, so a part treated on Monday and bonded on Thursday can show this exact failure even though the treatment step was technically completed.
Signature: Adhesive Residue Left on Both Faces After Separation
A joint that fails with epoxy visible and adhered to both plastic surfaces, rather than releasing cleanly from one, points to cohesive failure within the adhesive itself rather than a surface preparation problem. This usually means the epoxy was undercured — either short of its specified cure time, cured at too low a temperature, or mixed off-ratio in a way that left the cross-link density below spec. Because this failure signature can look superficially similar to a properly bonded joint at a glance, checking cured hardness against the data sheet value is a faster diagnostic step than assuming the mix ratio was correct.
Signature: The Plastic Itself Tears or Cracks at the Joint
This is the signature everyone is actually trying to achieve, not a failure at all — it confirms the adhesive bond exceeds the strength of the substrate material. If this signature shows up during a destructive pull test on a qualification sample, the joint is performing as intended and no further adjustment to adhesive selection or surface prep is needed.
Signature: Stress-Whitening or Crazing Radiating From the Bond Edge
Fine cracks or a whitened haze appearing in the plastic near the bond line, without the bond itself separating, points to chemical attack rather than a mechanical failure. Polycarbonate and acrylic in particular are prone to solvent stress-cracking when exposed to an incompatible cleaning solvent, an uncured epoxy amine that migrated into the plastic before full cure, or even certain cosmetic finishes applied near the joint after bonding. Because this reaction can take days to become visible, a joint that looks clean immediately after assembly can still develop this signature later — testing solvent compatibility on a scrap piece of the actual production plastic, not a generic reference sample, is the only reliable way to rule this out in advance.
Signature: A Bond That Warps or Bows the Assembly After Cure
Visible distortion in a large, flat bonded panel — even when the bond itself holds — traces back to cure shrinkage rather than adhesive failure. Two-part epoxies shrink roughly one to three percent by volume during cure, and on a large flat bonded area that shrinkage pulls the two plastic faces together unevenly, locking bending stress into the assembly before it ever sees a service load. A joint that shows this signature needs either a lower-shrinkage grade, a thinner and more uniform bond line, or, for genuinely large panels, a semi-flexible adhesive that relaxes shrinkage stress rather than storing it as a permanent bow.
Signature: Intermittent Failures Across an Otherwise Consistent Production Run
When most parts from a batch hold and a minority fail with no obvious pattern, the cause is frequently plastic identification rather than process control — a supplier substituting a visually identical but chemically different resin into a batch of parts, undetected until bonding. A water-float check separates polyethylene and polypropylene from denser common plastics in seconds, and a solvent-spot test with acetone on a hidden area differentiates ABS and polystyrene (which soften) from polycarbonate and acrylic (which craze) from polyolefins and acetal (which show no reaction) — both quick enough to run on a sample from a suspect batch before committing it to production.
Building the Fix Into the Process, Not Just the Repair
Diagnosing a fracture signature tells you what went wrong on the parts you already have; preventing a recurrence means writing the fix into the standard process — a defined activation-to-bond time window for treated plastics, a hardness spot-check on cured coupons pulled from each shift, and solvent-compatibility testing on any new cleaning agent or finish before it’s introduced near a bonded joint. Incure supplies two-part epoxies and UV-curable Uni-Weld Plastic Bonder grades and supports manufacturers through exactly this kind of failure-mode diagnosis.
For assemblies where a large flat bonded area shows shrinkage-driven distortion, the same underlying stress mechanism is covered from a thermal-cycling angle in how CTE mismatch causes adhesive bond failure — cure shrinkage and thermal-expansion mismatch both load a rigid bond line in ways a toughened grade absorbs better than a hard one.
For the foundational steps this diagnostic guide assumes — plastic identification, surface preparation, and joint design for a new bond rather than a failure already in hand — see a guide to plastic-to-plastic bonding with epoxy glue.
If you’re looking at a batch of failed parts and the fracture signature doesn’t clearly match one of the patterns above, Email Us with photos of the failure and the plastics involved — Incure’s applications team can often narrow the cause from the break pattern alone.
Contact Our Team to review a specific plastic-to-plastic bonding failure or to qualify a new joint design before it reaches production.
Visit www.incurelab.com for more information.