Quick answer. A footwear manufacturer should select polyurethane adhesive by outsole chemistry, upper material, line speed, activation equipment, and the required peel result after ageing—not by glue price per kilogram. For cemented construction, stable shoe sole bonding usually requires material-specific surface treatment, controlled adhesive coat weight, complete flash-off, heat reactivation, uniform pressing, and release only after the specified cure period. Validate the complete process on production materials because rubber compounds, EVA density, PU skin, plasticizer level, and mould-release residue can change adhesion even when the material name remains unchanged.
The purchasing decision therefore covers a bonding system: adhesive, primer or surface treatment, hardener where specified, application equipment, operating window, test method, and batch documentation. A low-priced drum can become the higher-cost choice if it adds one operator, extends the drying conveyor, or produces edge opening after flexing.
Map the Sole and Upper Before Selecting PU Glue
Start with the exact material pairing and construction. Record the outsole compound, hardness, density in kg/m³, colour, recycled-content percentage, mould-release agent, upper backing, bonding margin width in mm, and whether the sole is newly moulded or has been stored for several weeks.
“Rubber sole” is not a sufficient specification. Vulcanised SBR, SBS thermoplastic rubber, nitrile rubber, and natural-rubber-rich compounds present different surface chemistry. Research on shoe-soling materials found that PVC and PU could be bonded with polyurethane adhesive without chemical modification in the studied system, while SBS and SBR required chemical surface modification; see the International Journal of Adhesion and Adhesives study on solvent-free shoe bonding.
| Bonding surface | Main production risk | Trial preparation route | Evidence to record |
|---|---|---|---|
| Vulcanised SBR or rubber | Low wettability, wax bloom, sulphur compounds, or mould-release residue | Abrade without cutting through the skin; remove dust; apply the approved rubber treatment or validated plasma process | Abrasion tool and grit, treatment batch, application time, treated-to-coated interval |
| SBS or TPR outsole | Low-polarity surface and variation between compound recipes | Use a treatment proven on the exact compound; avoid transferring a successful SBR recipe without testing | Compound code, contact angle if available, peel result and failure mode |
| Moulded PU sole | Weak moulded skin, release agent, hydrolytic degradation, or incomplete cure | Clean or abrade according to the sole supplier's instruction; reject crumbly or chemically degraded material | Sole age, storage humidity, preparation method, cohesive versus interfacial failure |
| EVA midsole | Low surface energy, foam tearing, or excessive adhesive absorption after aggressive roughing | Light controlled roughing plus the approved EVA primer, corona, plasma, or other validated treatment | EVA density in kg/m³, hardness, treatment setting, coat weight in g/m² |
| PVC sole | Plasticizer migration and surface contamination | Use a PVC-compatible PU system; confirm ageing performance rather than relying on initial peel | PVC formulation or supplier code, 24-hour peel, heat-aged peel, bond-line condition |
| Leather upper margin | Finish coating blocks penetration, or heavy roughing weakens the leather | Remove the finish only inside the marked bonding margin and brush away fibres and dust | Roughing depth, margin width in mm, leather tear or adhesive failure |
| PU-coated synthetic or textile | Coating delamination, heat distortion, or adhesive strike-through | Test the coating-to-backing strength first; reduce heat and coat weight if the face deforms | Upper construction, reactivation temperature, appearance, failure layer |
Surface treatment has an operating lifetime. A treated rubber sole left through a shift change may collect dust or lose surface activity. Establish a maximum interval in minutes by testing parts bonded immediately, after 30 minutes, after 60 minutes, and at the proposed production limit.
Do not use a water-break appearance, fingertip tack, or operator judgement as the sole release criterion. Record the actual peel force and the failure location: adhesive-to-sole failure, adhesive-to-upper failure, cohesive split inside the glue, foam tear, rubber tear, or coating delamination.
Compare PU Adhesive Systems Against the Production Line
A polyurethane adhesive shoe sole bonding footwear manufacturer normally evaluates four system families. The values below are engineering trial windows commonly encountered in footwear processing, not a substitute for the selected supplier's technical data sheet. Confirm temperature, coat weight, open time, mix ratio, and cure time for every approved formulation.
| Selection dimension | 1K solvent-based PU | 2K solvent-based PU | Water-based PU | Reactive PUR hot melt |
|---|---|---|---|---|
| Physical form | Ready-to-use solution | Base adhesive plus specified isocyanate hardener | PU dispersion in water; some systems use a crosslinker | Moisture-reactive, 100% solids melt |
| Typical trial condition | Wet coat about 80–150 g/m²; flash and reactivate around 50–70°C | Similar application route, with supplier-defined mixing and limited pot life | Dry with controlled airflow and roughly 50–70°C heat before assembly | Apply molten, commonly within an equipment window near 100–140°C |
| Line constraint | Solvent flash-off, local exhaust, fire control, and reactivation stability | All 1K controls plus mix accuracy, clocked pot life, and container identification | Water removal can become the bottleneck in humid factories | Heated tank, hose, nozzle, moisture exclusion, and purge discipline |
| Handling strength | Develops after correct flash, reactivation, and pressure | Good early handling with additional chemical crosslinking | Weak if assembled before water has left the film | Green strength develops within seconds as the melt cools; chemical cure continues afterward |
| Common failure | Solvent trapped under a dry skin or adhesive cooled before pressing | Wrong hardener dose, expired pot life, or gel contamination | Milky wet film, foam moisture, slow drying, or poor wetting | Charred material, nozzle blockage, short open time, or moisture-damaged stock |
| Best procurement fit | Existing cementing lines with validated exhaust and heat activation | Footwear requiring higher heat or ageing resistance after validation | Plants reducing solvent use that can supply enough drying capacity | Automated or single-side processes designed around molten application |
The table prevents a false comparison based only on quoted USD per kg. Calculate adhesive cost per pair from solids delivered to the bond line: wet coat weight multiplied by coated area and application sides. Then add primer, hardener, cleaning solvent, oven energy, scrap, exhaust maintenance, and labour.
For example, a double-side process using 120 g/m² wet on a total coated area of 0.040 m² consumes about 4.8 g of mixed adhesive per pair before transfer loss. At 15% application loss, purchasing consumption becomes about 5.5 g per pair. A USD 0.50/kg quotation difference changes material cost by only USD 0.0028 per pair at that consumption; one additional rework operation can outweigh it.
Reactive PUR can remove solvent flash-off and may permit one-side coating, but the plant must control melt temperature and residence time. Puglue describes the intended footwear workflow on its single-side PUR footwear adhesive page. Treat any published temperature or performance range as a screening point until trials use the buyer's outsole, upper, press, and climate.
Control the Bonding Process From Roughing to Cure
1. Prepare a clean, repeatable bonding margin
Mark the cementing margin with a template. Rough only the area that will be covered; an exposed rough line is a cosmetic defect, while an unroughed strip near the feather edge becomes a peel initiation point.
Replace worn abrasive wheels by a defined trigger such as pairs processed or measured wheel diameter. Vacuum extraction should capture dust at the tool. Compressed-air blowing can move rubber dust onto prepared uppers and nearby adhesive films.
2. Apply primer and adhesive by measured coat weight
Check viscosity using the method and temperature named in the supplier's specification. A viscosity result without cup type, spindle, speed, or test temperature cannot be compared between incoming lots.
Run a gravimetric check at start-up and after lunch: weigh a test piece, coat a known area, allow the prescribed volatile loss if required, and reweigh. Report g/m² rather than “thin” or “heavy.” A practical development matrix can test 80, 110, and 140 g/m² wet while holding every other variable constant.
Too little adhesive leaves dry islands at leather fibres and sole texture peaks. Too much can trap solvent or water, lengthen drying, create a soft bond line, bleed beyond the margin, and raise consumption without increasing peel force.
3. Control flash-off and reactivation
Flash time must be stated with ambient temperature and relative humidity. Ten minutes at 30°C and 45% RH does not represent ten minutes at 18°C and 85% RH, particularly for a water-based dispersion.
Measure the bond-line surface temperature with a verified infrared instrument or contact sensor immediately after reactivation. Oven setpoint is not part temperature. Sole colour, mass, distance from the heater, lamp age, belt speed, and loading density all change the temperature reaching the adhesive film.
Build a trial around the supplier's window—for example, 50°C, 60°C, and 70°C at the bond line—then inspect heat-sensitive foam, synthetic coating, and decorative finishes. Do not adopt these example levels if the adhesive data sheet specifies another range.
4. Assemble and press before the activation window closes
Track seconds from oven exit to first pressure. The heel seat, toe, waist, and sidewall must contact before the film cools below its effective activation range. A queue between the heater and press can produce acceptable flat-panel tests yet fail at the toe edge.
Specify pressure in MPa or controlled machine pressure plus fixture area, not “press firmly.” Establish the line setting with pressure-indicating film or a mapped fixture check because nominal pneumatic pressure does not prove uniform pressure across a curved sole.
A pilot plan can compare 0.3, 0.5, and 0.7 MPa for 8, 12, and 15 seconds, within the limits of the shoe structure and equipment. Excess pressure can collapse EVA cells, distort the foxing, or squeeze adhesive away from the edge; insufficient pressure leaves voids and poor molecular contact.
5. Cure, condition, and test by lot
Do not flex-test or pack a chemically curing bond immediately after pressing. Define minimum rack time, commonly evaluated at 24, 48, and 72 hours during qualification, and confirm the approved release time from measured results.
ISO 17708:2018 describes upper-to-sole adhesion testing for whole footwear and includes ageing conditions usable for production control. ISO 18454:2018 defines standard atmospheres for conditioning and testing footwear components, so the laboratory should state which atmosphere it used.
For rubber-to-flexible-substrate development, ASTM D413-98(2025) measures the force per unit width needed to separate a rubber layer from a flexible substrate. It does not supply one universal footwear pass value; the buyer must set the acceptance threshold for the shoe category, construction, size, location, and ageing condition.
A useful control record contains force in N/mm, specimen width in mm, test speed in mm/min, conditioning time, ageing exposure, and failure mode. An average of 3.0 N/mm can hide five results of 4.0, 3.8, 3.6, 2.1, and 1.5 N/mm; therefore set both an average and an individual minimum.
Write a Purchase Specification That a Manufacturer Can Hold
Manufacturer-direct sourcing works best when the inquiry contains production evidence. Send bonded material samples or clearly identified sole and upper pieces, current line speed in pairs/hour, application method, oven length, measured bond-line temperature, press setting, target open time, destination climate, and the buyer's peel protocol.
Ask the adhesive manufacturer to return a written recommendation covering surface treatment, application sides, coat-weight range, flash window, activation range, assembly window, pressing condition, cure time, storage temperature, unopened shelf life, and incompatibilities. “Suitable for shoes” is not an approval specification.
For customization, change one formulation target at a time. Examples include extending assembly time from 90 to 180 seconds, lowering activation temperature by 10°C for a heat-sensitive upper, increasing initial grab for a high sidewall, or resisting plasticizer migration from a defined PVC compound. Each change must be tested against peel, flexing, heat ageing, hydrolysis exposure where relevant, colour, odour, viscosity stability, and application behaviour.
Freeze the approved sample with a supplier code, revision, manufacture date, and retained reference. Incoming quality control should compare appearance, solids or non-volatile content, viscosity using the stated method, density where specified, and a standard bond panel. A certificate of analysis should show actual batch values and limits rather than a repeated “pass.”
Request the current technical data sheet, safety data sheet, transport classification, country-of-origin documentation, and declarations required by the destination market. If a supplier presents ISO 9001 or another management-system certificate, verify the certificate number, validity date, issuing body, manufacturing-site address, and scope; a logo on a brochure does not prove that the adhesive plant falls within the certified scope.
For EU industrial or professional use, examine the free-diisocyanate content and applicable training obligations. REACH Annex XVII Entry 74 addresses mixtures containing at least 0.1% by weight total diisocyanates and specifies training elements for covered uses. Confirm the current legal duties with the importer and safety specialist rather than treating a low-VOC claim as proof of worker protection.
Engineering controls remain necessary where exposure is possible. The NIOSH isocyanate guidance identifies closed systems and ventilation as primary exposure controls and explains that inhalation or skin exposure can cause sensitisation. Review SDS-defined gloves, local exhaust, storage, spill response, and waste handling before introducing a hardener or reactive hot melt.
Commercial approval should follow a staged order: laboratory panels, 20–50 trial pairs, one controlled production shift, and then a monitored first batch. Agree the MOQ in kg or tonnes, packaging size, production lead time in calendar days, retained-sample period, batch traceability, change-notification period, complaint-sample procedure, and replacement terms before the first container ships.
FAQ
Q: Which polyurethane adhesive is best for shoe sole bonding?
The correct choice depends on the exact upper–sole pair and production line. Solvent-based PU suits many established heat-reactivation lines; water-based PU requires enough drying capacity; 2K PU adds a controlled hardener step; reactive PUR hot melt suits lines equipped for heated application and moisture-controlled storage. Select by aged peel and failure mode, not initial hand pull.
Q: Why does the sole peel even though the adhesive feels tacky?
Tack measures surface stickiness, not a completed bond. Common causes are mould-release contamination, untreated SBR or SBS, adhesive assembled while still wet, reactivation below the required bond-line temperature, delay before pressing, uneven pressure, an expired 2K mixture, or testing before chemical cure finishes.
Q: Does EVA always need primer before PU glue?
No universal rule covers every EVA compound. Density, vinyl-acetate content, foam skin, recycled content, release agent, and storage history affect wetting and foam strength. Test untreated, lightly abraded, and supplier-approved treatment routes on the production EVA, then compare N/mm results and check whether failure moves from the interface into the foam.
Q: What peel strength should a footwear factory require?
ISO 17708 defines a whole-shoe test method but does not give one pass value for every shoe. The brand or factory should specify test location, specimen width, conditioning, ageing, test speed, average N/mm, individual minimum, and acceptable failure mode for each construction. Safety footwear, fashion shoes, sandals, and flexible sports shoes should not inherit one arbitrary threshold.
Q: Can a factory switch from solvent-based PU to water-based PU without changing the line?
Usually not without a capacity study. Water removal depends on coat weight, airflow, temperature, humidity, substrate absorption, and conveyor residence time. Run a mass-loss or dryness study at peak seasonal humidity and confirm that the oven can dry every coated zone without shrinking EVA or deforming the upper.
Q: How should purchasing compare adhesive quotations?
Convert each offer into cost per accepted pair. Include mixed adhesive consumption in g/pair, primer and hardener, transfer loss, labour, drying energy, ventilation, cleaning, scrap, rework, packaging size, MOQ, shelf-life loss, freight, and lead time. Require the compared products to pass the same aged-peel protocol before comparing commercial totals.