Industrial adhesive specification is a four-axis decision: substrate pair, mechanical load, chemical/thermal environment, and process throughput — and every qualified supplier, including DELO and the Omnia Koll portfolio of Jowat, Kömmerling and Arkema-Simson, frames its datasheet around those four axes [S3][S6].
Specifiers who collapse the decision to "strength number" routinely over-engineer cost or, worse, under-spec a bond that fails in service. The five structural-adhesive families — epoxy, polyurethane, methacrylate/acrylic, silicone, and hot-melt — cover roughly 90% of factory and field bonding work, and each carries a distinct fingerprint on cure chemistry, gap tolerance, and temperature ceiling. Picking the wrong family is the single most common root cause of bonded-assembly warranty claims; getting the family right is a 70% problem, with surface prep, joint design, and cure schedule settling the rest.
Substrate Pair and Surface Energy: The First Gate
Adhesive selection begins with the substrate surface energy, not the adhesive datasheet — metals typically sit at 30–70 dyn/cm and accept most structural chemistries, while polyolefins (PP, PE) drop below 30 dyn/cm and force the specifier toward surface pretreatment (corona, plasma, flame) or toward hot-melt and specialty acrylic systems [S3]. DELO positions its bonding portfolio explicitly around "material diversity" and miniaturisation, a direct acknowledgement that mixed-substrate micro-joining is now the rule, not the exception, in electronics and medical assembly [S3].
For dissimilar pairs — aluminium-to-CFRP, glass-to-PA, rubber-to-metal — the coefficient of thermal expansion mismatch and the relative modulus usually drive the call more than the quoted lap-shear strength. Cyanotec's ISO 9001:2015-certified technical service workflow starts every enquiry with a substrate-bonding questionnaire before any chemistry is proposed, which mirrors the discipline most quality systems now require [S1]. Joint design then has to compensate: thicker bondlines (0.1–0.5 mm) for CTE stress, thinner bondlines (≤0.05 mm) for thermal-path and optical clarity.
Chemistry Family Comparison: Five Options Against Four Decision Criteria
The five dominant structural and assembly adhesive families line up against four decision criteria as follows, with values reflecting the operating envelopes published by DELO and by the Jowat/Kömmerling/Arkema-Simson lines carried by Omnia Koll [S3][S6]:
Two-part epoxy typically delivers 15–40 MPa lap-shear on metals, continuous service from roughly -55 °C to +150 °C (specialty grades to 200 °C), excellent chemical resistance, and a fixture time of minutes to hours depending on hardener. Polyurethane bonds flex with substrates, tolerates differential thermal expansion well, runs from about -40 °C to +80 °C continuous, and is the workhorse for automotive composite and film-laminate work where Jowat and similar furniture-industry supply lines are heavily used [S6]. Methacrylate/acrylic (often called MMA) accepts minimal surface prep, fills large gaps up to several millimetres, and tolerates oily metals — useful in field repair and truck-body assembly. Silicone seals and bonds in continuous service from -60 °C to +200 °C and beyond, but with lower cohesive strength (typically 1–3 MPa), so it is rarely the structural choice. Hot-melt (EVA, polyamide, PUR-reactive) applies in seconds with no cure wait, suits high-throughput packaging and product assembly, and is the dominant format across the industrial-adhesives and labelling value chain that HERMA serves for rugged industrial labels [S2][S6].
Reactive PUR hot-melts blur the line: they apply as a melt, then crosslink with moisture for a final thermoset bond with much higher heat resistance than a standard EVA hot-melt — a key reason PUR hot-melt is replacing solvent-based contact adhesives in furniture and mattress lines across the Jowat customer base [S6]. Specifiers comparing families should score each option against the same four criteria — substrate, load, environment, process — and reject any single-axis "winner" decision.
Cure Mechanism, Open Time, and Process Throughput
Cure behaviour drives line throughput more than datasheet strength does. Two-part epoxies offer a defined mix ratio and predictable fixture (5–60 min at room temperature) but require meter-mix dispensing capital; one-part heat-cure epoxies reach full strength only above roughly 120–180 °C, ruling them out for temperature-sensitive substrates [S3]. UV/light-cure acrylics — a DELO specialty — fixture in 1–5 seconds under the right lamp intensity, which collapses cycle time in medical, electronics, and glass-bonding cells [S3].
Moisture-cure silicones and polyurethanes skin over in minutes but continue to cure inward for 24+ hours, so jigs and fixturing must hold the bondline through the full depth-of-cure window. Hot-melts and reactive PUR hot-melts eliminate the cure-wait entirely but introduce a pot-life constraint on the melt tank (typically 4–8 hours before char and viscosity drift degrade bond quality) [S6]. When cycle-time is the binding constraint, a slower-curing but stronger structural adhesive is often the wrong answer, and a UV-cure or hot-melt system paired with a redesigned joint will outperform it on total cost-per-bond.
Environmental Exposure: Temperature, Chemical, and Outgassing
Three environmental axes disqualify adhesive choices faster than any strength number. Continuous service temperature: silicones reach +200 °C and beyond where epoxies soften, while flexible PU drops off above +80 °C wet. Chemical exposure: epoxies resist fuels, oils, and many solvents; silicone resists water and weathering but is attacked by some fuels; polyurethane sits in the middle and is the default for fuel-resistant bonding in marine applications where Kömmerling's marine-industry line is positioned [S6]. Outgassing and cleanliness: low-outgassing grades (often meeting NASA ASTM E595 equivalents) are mandatory in vacuum, optics, and semiconductor applications — a non-negotiable for any bonded component going into a process tool.
For labelling and lightweight industrial identification — a parallel use case for adhesives — HERMA emphasises that "labels for industrial applications are an economical means of ensuring reliable processes," with adhesive chemistry tuned to surface, temperature, and chemical exposure of the host asset rather than raw adhesion number [S2]. The same logic applies to load-bearing bonds: pick the chemistry whose environmental fingerprint matches the operating envelope, then validate with a representative coupon test.
Standards, Certification, and Quality-System Gates
Adhesive specification rarely stands alone: the supplier's quality system and the application-specific certifications carry equal weight. ISO 9001:2015 is the baseline; Cyanotec, like most serious industrial-adhesive suppliers, publishes this certification as a contact-page prerequisite and ties technical advice to that audited QMS [S1]. Application-specific standards then layer on top — automotive specs (OEM-specific), aerospace (often BMS, DMS, or MIL-HDBK-17 references), food contact (FDA 21 CFR / EU 1935/2004 frames), potable water (WRAS, NSF 61), and electronics (UL 746C for polymeric adhesives).
For buyers, the practical gate is to demand the actual test report, not a generic compliance letter. A datasheet claim of "meets FDA" means little without the specific food-type exposure and temperature range tested, and a UL listing is meaningless without the correct end-use configuration. Cross-reference the certificate number on the issuing body's database before releasing the part to production — a step the LookChem-traded distributor Industrial Adhesive Enterprises (Mumbai, India) and similar trading channels often cannot provide, which is one reason direct-OEM sourcing dominates for safety-critical bonds [S7].
Joint Design and Surface Preparation: The Variables That Override Datasheets
A correctly designed and prepared joint reaches 60–90% of the substrate's intrinsic strength regardless of adhesive chemistry; a poorly designed joint fails at 10–30% of the rated value on the best adhesive. The rules are mechanical, not chemical: lap-shear samples should be designed so the adherend yields before the bondline; overlap length should be tuned to substrate thickness (the classic 10–20× adherend-thickness rule for thin metal joints); and peel loads should be avoided in favour of shear, tension, or compression paths wherever geometry permits. [S1]
Surface preparation typically drives more variance than adhesive choice. Abrasion + solvent wipe, grit-blast + silane treatment, or plasma/corona treatment for polyolefins can each shift bond strength by 2–5× on the same adhesive. For structural work, document and audit the prep sequence; for high-volume production, integrate the prep into a single in-line cell so the operator cannot skip a step.
Cost-per-Bond and Total Cost of Ownership
Adhesive price per kilogram is a poor decision metric: a 5× more expensive adhesive that cures in 5 seconds and requires no fixturing almost always wins on total cost-per-bond against a cheaper adhesive that needs jigs, ovens, and floor space. A reasonable cost-per-bond model should include adhesive mass per assembly, dispensing and metering cost, fixturing and floor-space cost, rework and scrap rate, and warranty exposure. PUR hot-melt lines replacing solvent-borne contact adhesives in furniture manufacturing — a documented shift across Jowat's customer base — usually pay back the equipment premium inside 12–24 months on solvent abatement, ventilation, and floor-space savings alone [S6].
For lower-volume or higher-mix operations, meter-mix dispensing of two-part epoxies, or bench-top UV cure cells running DELO-style light-cure acrylics, keep capital low and qualification fast, with the trade-off being a smaller chemistry window [S3]. Industrial traders such as the Mumbai-based Industrial Adhesive Enterprises position around that flexibility, supplying small-batch glue and binder products to customers who cannot justify a direct-OEM relationship [S7]. Whatever the supply model, the evaluation unit should be the cost-per-bonded-assembly, not the line-item adhesive cost.
Three signals are worth tracking through the rest of 2026: the continued displacement of solvent-borne contact adhesives by reactive PUR hot-melt in furniture and mattress lines [S6]; the broadening of light-cure acrylics into higher-temperature and dark-substrate applications as LED-UV lamp intensity climbs [S3]; and the tightening of low-outgassing and traceability documentation across electronics, semiconductor, and medical-device supply chains — a documentation load that increasingly disqualifies trading-channel sourcing [S7]. A specifier who builds the selection matrix around substrate, load, environment, and process from the first enquiry will land a bond that survives service; a specifier who starts from a strength number is buying rework.
For component-level specifications, see industrial adhesive, industrial borescope, and industrial buzzer.