Quick answer. Polyurethane adhesive bubbles form when moisture reacts with free isocyanate and releases carbon dioxide, or when air, solvent, and volatile contaminants become trapped before a thick glue line can vent. Stop the defect by measuring component and substrate moisture, controlling the two-component mix ratio, reducing air entrainment, and validating cure behavior at the actual bond thickness. Do not solve recurring foam by increasing clamp pressure alone; pressure can hide surface bubbles while leaving internal voids that reduce bonded area.
A thick bond line changes the failure mechanism. A 0.10 mm film may release gas through an edge or surface, while a 2-5 mm layer can skin over and trap gas inside the joint.
The troubleshooting sequence below applies to one-component moisture-curing PU, two-component polyurethane adhesive, solvent-borne PU, and moisture-reactive PUR hot melt. The probable cause and corrective action differ among those four chemistries.
Why Thick Polyurethane Glue Lines Bubble or Foam
Water creates carbon dioxide inside isocyanate-containing adhesive
Water reacts with an isocyanate group to form an unstable carbamic acid, which decomposes into an amine and carbon dioxide. The amine can then react with another isocyanate group to form a polyurea structure, so moisture changes both the cell structure and the cured polymer chemistry.
The gas volume is larger than many production teams expect. Under an ideal-gas calculation at about 25°C and 1 atmosphere, 1 g of water can generate approximately 1.36 L of carbon dioxide if the reaction proceeds completely.
A 100 mm × 100 mm joint with a 2 mm gap contains only 20 mL of adhesive. Just 0.10 g of reactive water has a theoretical gas yield near 136 mL, more than six times that joint volume.
Moisture can enter through hygroscopic polyols, mineral fillers, wooden panels, concrete, fabric, leather, paper honeycomb, humid compressed air, a washed mixing vessel, or packaging left open. Condensation also forms when a cold drum is opened in a warm, humid plant.
Air, solvent, and temperature can produce similar-looking voids
Fast mixing can fold air into a viscous polyol or filled adhesive. A planetary blade that forms a deep vortex, a leaking pump seal, or a static mixer operated below its designed flow range can introduce dispersed bubbles before dispensing.
Solvent-borne shoe and lamination adhesives can blister when the assembly is closed before the solvent has flashed. Water-based primers, cleaning alcohol, plasticizer, release agent, and low-boiling substrate contamination can create the same symptom.
Temperature adds two effects. Heating lowers viscosity and may improve wetting, but it also accelerates reaction and expands existing gas; cooling raises viscosity and makes entrained air harder to release. A component stored at 10°C and applied without conditioning can therefore behave differently from the same batch at 25°C.
Each PU adhesive family has a different risk profile
- One-component moisture-curing PU: Water is part of the cure mechanism, but excess water can create visible foam. A thick section may develop a cured surface skin while its center remains soft or gas-filled.
- Two-component PU: The intended polyol-isocyanate reaction does not require gas generation. Moisture contamination, incorrect A:B ratio, poor mixing, or filler settling is therefore more suspicious than ambient cure humidity alone.
- Solvent-borne PU: Residual solvent becomes the first suspect when bubbles appear after mating, heat activation, or oven exposure. Measure flash time and retained mass instead of assuming every bubble is caused by water.
- Reactive PUR hot melt: Moisture participates after application, but excessive substrate moisture, overheated melt, long residence time, or degraded material can cause porosity and viscosity drift.
Read the Defect Before Changing the Formula
Cut at least five joints through the center and near both edges. Record bubble diameter, depth, distribution, cure state, and the side to which each void is attached; this pattern often separates a wet substrate from mixing or dispensing faults.
| Observed defect | Most probable mechanism | Confirmation test | First containment action |
|---|---|---|---|
| Fine cells throughout the full thickness | Moisture distributed in a component or filler; air incorporated during mixing | Karl Fischer moisture test on both components; cure a hand-mixed sample without the production pump | Quarantine the open container, dry approved fillers, and compare low-shear mixing with production dispensing |
| Large voids attached to one substrate | Wet or porous substrate releasing vapor or air | Measure substrate mass before and after a supplier-approved drying cycle; coat each substrate separately | Dry and condition the substrate, then seal porous surfaces if the application permits |
| Bubbles concentrated near the center of a 2-5 mm line | Gas trapped after the surface or edges gel | Prepare stepped gaps of 0.5, 1, 2, 3, and 5 mm at identical temperature | Reduce the gap, slow the cure, provide an escape path, or specify a formulation validated at the required thickness |
| Blisters after heat activation or oven cure | Residual solvent, water, or another volatile expanding under heat | Compare coated-part mass before coating, before mating, and after drying; run a slower temperature ramp | Increase flash time or staged drying without exceeding the substrate or adhesive temperature limit |
| Intermittent foam after pump start-up | Air pocket, leaking suction connection, empty drum, or an unmixed static-mixer front | Collect sequential 50-100 g shots and compare density and A:B mass ratio | Purge the line, repair the leak, and discard a validated start-up volume |
| Soft zones beside hard or brittle zones | Off-ratio two-component dosing or incomplete mixing | Catch A and B separately for 30-60 seconds and weigh them; inspect mixer elements | Recalibrate pumps and replace the mixer before changing catalyst or cure temperature |
| Foam only late in a shift | Open-time moisture pickup, pot-life overrun, temperature rise, or PUR residence-time degradation | Plot viscosity, component temperature, shot density, and elapsed open time every 30 minutes | Use smaller day tanks, shorten exposure, and define maximum open-container or pot-life time |
Density is a useful screening measurement. Fill a calibrated cup without trapping new air, level it, and divide net mass by cup volume; compare the result with an accepted unopened batch at the same temperature.
A density loss indicates gas or air, but it does not identify the source. Combine density with Karl Fischer moisture, gravimetric mix-ratio checks, viscosity, and cross-section microscopy or a low-magnification image with a scale.
Build a Controlled Process Window
Measure moisture instead of relying on room humidity
Relative humidity does not reveal how much water is stored inside wood, filler, fabric, or a polyol blend. Specify a moisture test method and sampling point for each material that can contact the isocyanate.
ASTM D6869 covers Karl Fischer determination of moisture in plastics, and ISO 760 describes Karl Fischer determination of water. The laboratory must verify that the sample dissolves or releases its water under the chosen extraction conditions; a number from an unsuitable method can look precise while missing bound water.
For a plant trial, record adhesive moisture in ppm or %, substrate moisture in %, ambient temperature in °C, relative humidity in %, and compressed-air pressure and dew point. A compressed-air pressure reading alone cannot prove that the air is dry.
As an engineering starting point, dry process air with a pressure dew point at or below -20°C is far less likely to introduce water than untreated shop air. The final limit must come from trials with the actual formulation, equipment, and substrate.
Control storage, conditioning, and container exposure
- Condition sealed drums to the supplier-approved processing range before opening. If a drum arrives cold, allow its contents and headspace to equilibrate while sealed so humid air cannot condense inside.
- Use dry nitrogen blanketing or a desiccant breather for moisture-sensitive day tanks where the equipment design and safety assessment permit it.
- Record the first-open date and cumulative open time. A vague instruction such as “keep dry” cannot be audited after a defect.
- Take samples through a dry valve into a sealed container. An open cup carried across a humid shop may no longer represent the drum.
- Keep water-based primers and freshly washed tools physically separated from isocyanate-side transfer equipment.
Verify ratio, mixing, and dispensing
For two-component PU, check the specified ratio on the correct basis. A formulation stated as 100:25 by mass is not interchangeable with 100:25 by volume unless component densities make those quantities equivalent.
Catch each component separately, weigh it on a calibrated balance, and calculate the delivered ratio over at least three shots. A practical trial target can be within ±1% of the specified mass ratio, but the adhesive supplier’s allowable tolerance governs production release.
Inspect static mixers for the correct element count, internal diameter, and orientation. Low flow can leave striations, while excessive back pressure can distort the ratio or damage seals.
For batch mixing, keep the blade below the liquid surface and avoid a vortex that reaches the shaft. If vacuum deaeration is permitted, establish the pressure and time by trial; an aggressive vacuum can expand dissolved gas, strip solvent, or trigger moisture-curing material rather than correct it.
Design the joint so gas has somewhere to go
Use calibrated glass beads, wire spacers, shims, or molded stops to hold the intended gap. Reported adhesive consumption cannot substitute for direct thickness control because squeeze-out and substrate absorption change the final line.
Apply adhesive in a pattern that pushes air toward an open edge as the parts close. Dropping a large panel flat onto a complete wet film can trap pockets even when the adhesive itself is bubble-free.
Clamp pressure must distribute the adhesive without starving the joint. Verify the cured thickness at several points; a bond designed for 2.0 mm but crushed to 0.2 mm at the edges has two different cure conditions in one assembly.
Run Thick Bond Line Troubleshooting as a Designed Trial
Change one controlled factor at a time during initial diagnosis. Randomly adjusting moisture, catalyst, temperature, clamp pressure, and mix speed in the same run makes the result impossible to transfer to another line or batch.
Start with a 2 × 2 matrix: accepted versus suspect substrate, crossed with accepted versus suspect adhesive. Keep gap, applied mass, temperature, mixing method, and cure time constant.
If only the suspect substrate foams with both adhesive batches, focus on substrate moisture, porosity, primer, and contamination. If the suspect adhesive foams on both substrates, investigate component moisture, storage exposure, filler settling, and production mixing.
Next prepare stepped-gap coupons at 0.5, 1, 2, 3, and 5 mm, or across the actual drawing tolerance. Cure them at the lowest and highest plant temperatures expected during production.
Section each coupon after the specified cure time and calculate void area as a percentage of the bonded cross-section. Define acceptance numerically, for example “no void larger than 1.0 mm and total visible void area below 2%,” only after confirming that this threshold meets the mechanical and sealing requirements.
Test adhesion on conditioned, hot, cold, and humidity-aged specimens. ASTM D1002 or ISO 4587 can support lap-shear testing of rigid adherends, while ASTM D1876 or ISO 11339 can support T-peel work on suitable flexible assemblies; the joint geometry and failure mode must match the purchased application.
Record cohesive, adhesive, and substrate failure percentages. A nominal 5 MPa lap-shear value has a different meaning if the bond fails cohesively through the adhesive than if 80% of the area releases cleanly from an oily substrate.
For a rapid plant check, retain mixed adhesive from the beginning, middle, and end of a run in transparent cups of equal mass. Compare exotherm, gel time, cured density, cell structure, and hardness after the same conditioning period.
Specify a Bubble-Controlled PU Adhesive Purchase
A purchasing specification should state the substrate pair, primer, surface preparation, target gap and tolerance, application temperature, open time, press time, cure schedule, and required failure mode. “PU glue for metal” leaves the manufacturer unable to distinguish a 0.1 mm sheet lamination from a 4 mm battery enclosure gap.
For direct manufacturer development, send representative substrates rather than generic material names. Powder-coated steel, anodized aluminum, plasticized PVC, oily leather, and 12% moisture wood each require different wetting, moisture, and cure checks.
Request customization against measurable targets: mixed viscosity at a stated temperature and spindle method, non-sag thickness in mm, working time in minutes, fixture time in minutes, density in g/cm³, target cured hardness, and strength after the specified aging cycle. ASTM D1084 can be referenced for adhesive viscosity when its method suits the product.
Ask the manufacturer to run the same 0.5-5 mm stepped-gap test used by your plant. A sample that looks clear in a 0.2 mm drawdown does not establish that it will cure without foam in a 3 mm production cavity.
For each shipment, define the documents required: product data sheet, safety data sheet, batch number, manufacturing date, shelf-life statement, and certificate of analysis with the agreed release properties. If ISO 9001, REACH documentation, or another certificate is required, verify the certificate scope, issuing body, current validity, and the exact legal entity supplying the material.
European buyers should review the ECHA information on diisocyanates, including the REACH restriction and training requirements for industrial and professional use of products containing 0.1% or more monomeric diisocyanates by weight. Compliance documents do not replace workplace exposure controls.
Isocyanates can affect workers through inhalation and skin exposure. Process design, ventilation, personal protective equipment, spill procedures, and medical controls should follow the SDS and applicable law; the OSHA isocyanates guidance provides an authoritative starting point for US workplaces.
Commercial terms should be numeric as well. Put sample quantity in kg, pilot quantity in kg, production MOQ in tonnes, quoted lead time in calendar days, packaging size in kg, and retest or complaint response time in working days on the purchase record.
Manufacturer-direct supply is most useful when formulation changes are controlled. Require written notification before changes to isocyanate type, polyol package, filler, catalyst, solvent, production site, or test method, followed by requalification at the maximum bond thickness.
For a formulation discussion covering PU, lamination, hot-melt, and footwear bonding, use the polyurethane adhesive product category and submit the substrate pair, required glue-line thickness, monthly consumption, processing temperature, and destination market.
FAQ
Q: Why does a polyurethane adhesive foam only in a thick glue line?
Gas generated or introduced inside a thin film can escape before gelation. In a 2-5 mm section, the surface and edges may cure first, trapping carbon dioxide, air, or solvent in the center while reaction heat expands the bubbles.
Q: Can high humidity help a one-component PU cure without causing bubbles?
Moderate ambient moisture can accelerate surface cure, but relative humidity alone does not define the water entering the joint. Excessively wet wood, concrete, fabric, or filler can deliver far more water than the adhesive needs and create carbon dioxide faster than it can escape.
Q: What is the fastest way to distinguish moisture foam from entrained air?
Compare Karl Fischer moisture results with mixed-density and hand-mix tests. Foam that persists in a slow, dry hand mix points toward component or substrate moisture, while foam that disappears outside the production mixer points toward vortexing, pump leakage, or dispensing conditions.
Q: Will vacuum degassing eliminate polyurethane adhesive bubbles?
It can remove entrained air from some nonreactive or properly controlled components, but it cannot reverse carbon dioxide formation after water has reacted with isocyanate. Vacuum may also expand gas, remove solvent, or change a moisture-curing product, so the adhesive manufacturer must approve the pressure, temperature, and exposure time.
Q: Can extra clamp pressure stop PU foaming?
Pressure may compress bubbles while the joint is closed, but gas can re-expand after pressure release or remain as flattened internal voids. Measure cured thickness and cross-section void area rather than accepting a smooth outer bead as proof of a void-free bond.
Q: Should the buyer request a lower-viscosity adhesive for a thick joint?
Lower viscosity can improve wetting and air release, but it may increase sag, drainage, squeeze-out, and bond-line starvation. Specify viscosity, non-sag thickness, gap range, application orientation, and open time together, then validate all limits on production substrates.