Troubleshooting

Adhesive and Sealant Bonding Mistakes That Wreck Yield and How to Catch Them

The costliest adhesive and sealant failures trace back to a handful of repeatable errors in mixing, timing, dispensing, and cure. Here is how to spot each one and correct it with a number.

The most expensive mistake in two-part bonding is a mix ratio drift you cannot see. Symptom: parts pass initial handling but shear strength drops 30 to 50 percent after a week, and surfaces stay tacky. Root cause is a metering pump running off ratio, often a 2:1 epoxy dispensing closer to 2.6:1 as the resin side thickens in cold weather. The fix: verify ratio by weight, not volume, every shift. Shoot 30 grams of each side into cups and check against the Two-Part Mix Ratio calculator target within plus or minus 5 percent by mass. A 1.6:1 error on a 100:47 hardener already puts you outside spec.

Blown pot life shows up as a bond that never reaches full strength despite a correct ratio. Symptom: bubbles at the bond line, exotherm you can feel through gloves, and adhesive that gels in the static mixer mid job. Root cause is treating pot life as a fixed number when it halves for every 10 degree C rise. A material rated 45 minutes at 23 degrees C runs about 22 minutes at 33 degrees C. Fix: track working time against ambient with the Pot Life Usage calculator and size batches to consume material before 70 percent of the adjusted window, not the datasheet number.

Open time errors ruin bonds that looked perfectly mixed. Symptom: clean adhesive failure where the film peels off one substrate leaving no residue. Root cause is closing the joint after the open time window expired, so the adhesive skinned and lost wet out. On a moisture cure polyurethane, open time can be 8 to 12 minutes at 50 percent RH but collapses to 3 to 4 minutes at 80 percent RH. Fix: time from bead deposit to fixture, log it, and use the Open Time Window calculator to set a hard cutoff. If assembly takes longer, dispense in stages, not one long run.

Bond line thickness gets ignored until strength scatters wildly. Symptom: lap shear results swinging from 12 to 28 MPa on identical coupons. Root cause is a starved or flooded joint. Too thin, under about 0.1 mm, and you get adhesive starvation and stress risers. Too thick, over 0.5 mm on a structural epoxy, and cohesive strength falls as the glue line carries more load than it should. Fix: design for 0.15 to 0.30 mm using glass bead or wire spacers, and confirm with the Bond Line Thickness Control calculator. Measure squeeze out volume as a proxy on the line.

Bead volume math is a quiet unit trap. Symptom: you run short of adhesive halfway through a production batch you thought was fully costed. Root cause is confusing bead diameter with cross section. A round bead's volume scales with the square of diameter, so going from a 3 mm to a 4 mm bead is not 33 percent more material, it is 78 percent more. On a 1.2 meter perimeter that is roughly 15 cc versus 9 cc per part. Fix: compute from cross sectional area times length with the Adhesive Bead Volume calculator, then reconcile against actual cartridge counts weekly.

Cure time assumptions cause premature handling failures. Symptom: joints that pass at the workstation delaminate after the first thermal cycle or fixture removal. Root cause is confusing tack free time with functional cure. A cyanoacrylate is handleable in 30 seconds but reaches only 50 to 60 percent of ultimate strength for hours, and a heat cure epoxy needs the full time at temperature, not oven set point. Fix: base fixture release on the Adhesive Cure Time calculator using measured part temperature, and add a 20 percent margin for cold, thick, or high mass substrates that lag the oven.

Coverage and dispense rate mistakes compound scrap. Symptom: visible gaps in a sealant bead or over application that squeezes into threads and clearance holes. Root cause is a mismatch between robot travel speed and pump output, so a line rated 8 grams per second at 300 mm/s actually lays a thin bead at 400 mm/s. Fix: validate output against programmed speed with the Dispense Rate Achievement calculator and confirm real coverage against the Adhesive Coverage and Sealant Usage calculators. A 10 percent speed overshoot on a bead sized to tolerance leaves you 10 percent under filled.

Bad substrate data is the failure nobody logs. Symptom: intermittent bond failures on a part that ran fine for months, clustered on one supplier lot. Root cause is a surface energy or contamination change, a new mold release or an oil film, that no calculator catches because it is upstream of the math. Fix: dyne test incoming substrates to confirm at least 38 to 42 mN/m for good wet out on many plastics, and keep a retained sample per lot. When numbers look right but bonds fail, stop recalculating and check the surface first.

Published 2026-07-01.