Epoxy, acrylic (including MMA), cyanoacrylate, polyurethane, and silicone cover roughly 80% of general-fabrication bonding decisions, and most spec errors trace back to substrate mismatch or ignored cure kinetics rather than adhesive strength numbers [S8].
For general metal, plastic, and composite assembly, the working set narrows to five chemistries: two-part epoxy, structural acrylic/MMA, cyanoacrylate, polyurethane (PUR), and silicone, with hot-melt and pressure-sensitive tape as non-structural options for low-load or temporary joints [S1][S3][S8]. The right pick depends on substrate pair, required fixture time, service temperature, chemical exposure, and the need for flexibility or reworkability.
Structural chemistries: epoxy and acrylic (MMA) for load-bearing joints
Two-part epoxy is the benchmark structural adhesive for metal-to-metal and composite-to-metal joints, prized for high strength, temperature tolerance, and chemical resistance, with modern heat-accelerated formulations cutting cure times for aerospace and automotive chassis work [S1]. Modified epoxies and MS polymers extend that family into metal-fabrication shops that need slightly more flexibility on galvanized or coated steels [S2]. Acrylic adhesives, including methyl methacrylate (MMA) formulations, deliver a faster-curing alternative with strong adhesion to dissimilar materials, including plastics and coated metals, and resist impact and vibration in mixed-material vehicle and appliance assemblies [S1][S7]. For structural applications, epoxies and MMA adhesives are commonly specified as the first candidates, with silicone or polyurethane reserved for flexible bonds [S7]. The trade-off is well documented: two-part epoxies can need hours at room temperature to reach handling strength, while a typical MMA or modified acrylic reaches handling strength in 1-5 minutes at 20-25°C, which is why many high-mix fabricators keep both chemistries on the bench [S1][S6].
Fast-cure and flexible bonds: cyanoacrylate, polyurethane, silicone
Cyanoacrylates, marketed as instant adhesives, fixture in seconds on small plastic and metal components such as sensor housings and gaskets, but offer limited gap-filling and poor peel strength on flexible substrates [S1]. Polyurethane adhesives bridge the gap between structural rigidity and elastomeric flexibility, performing well on plastics, composites, and primed metals where differential thermal expansion is a concern [S2][S7]. Silicone adhesives and sealants sit at the upper end of the flexibility range, tolerating service temperatures typically from roughly -55°C up to 200-250°C depending on grade, and are favored where joint movement, UV exposure, or thermal cycling dominate the failure mode [S3][S8]. For vibration-prone enclosures or HVAC ducting in industrial coating shops, PUR and silicone outperform rigid epoxies because the adhesive must absorb shear rather than resist it. Practical rule from the bench: if the substrate pair has a coefficient-of-thermal-expansion mismatch above roughly 10×10⁻⁶/K, default to a flexible chemistry before re-evaluating epoxy with a primer.
Substrate matching: metal, plastic, glass, ceramic, rubber
Adhesion behavior is dominated by surface energy and surface roughness: mechanical interlocking dominates on rough metallic surfaces, while chemical bonding dominates on certain plastics [S1]. For thermoplastics, surface treatment (IPA wipe, abrasion, plasma, or chemical etch) often does more for joint strength than the chemistry choice itself, and a 2025 chemical-concepts guide specifically targets PMMA (acrylic) and PC (polycarbonate) bonding for signage, shields, retail displays, and protective barriers [S3]. Glass and ceramic bonding leans on UV/light-cure acrylates and modified epoxies because of their low-viscosity wetting on high-energy surfaces, while rubber-to-metal and rubber-to-plastic pairs are typically served by flexible PU or contact-cement systems because rigid epoxies crack under cyclic strain [S2][S3]. Designers selecting an industrial adhesive for mixed-substrate assemblies should map each pair to a recommended chemistry family rather than test a single adhesive across every joint, because cure kinetics, gap tolerance, and surface prep diverge sharply between pairs.
Selection framework: ASPEC and process-side constraints
3M's ASPEC method (Application, Substrate, Process, Environment, Cycles) is the most reproducible decision filter for general-fabrication shops, because it forces joint, surface, and manufacturing-process constraints onto the same checklist before chemistry is discussed [S5]. Process-side data is where most selections fail: two-part mix ratio, pot life, fixture time, full-cure time, viscosity at dispense temperature, and required bead profile all need to match the line takt before lab coupons are pulled [S5][S6]. A cyanoacrylate that fixtures in 10 seconds on a lab bench can take 30-60 seconds in a 15°C shop, so cure models must be validated at worst-case ambient, not room-condition spec sheets. For automated lines, viscosity stability over the dispense-pot lifetime and the ability to fixture through light oil contamination are the two most common gating parameters, and both belong in the Process leg of ASPEC rather than the Substrate leg [S5].
Comparison: five chemistries against four decision criteria
On the four criteria that drive most general-fabrication picks (cure speed, substrate range, flexibility, and chemical/temperature resistance), the five workhorse chemistries split cleanly. Two-part epoxy rates slow on cure, broad on substrate range, rigid on flexibility, and high on chemical/temperature resistance. MMA acrylic rates fast on cure, very broad on substrate range (including coated metals and many plastics), semi-rigid on flexibility, and medium-high on chemical/temperature resistance. Cyanoacrylate rates very fast on cure (seconds), narrow on substrate range (best on close-fitting plastics and metals with surface prep), rigid on flexibility, and low-medium on chemical/temperature resistance. Polyurethane rates medium on cure, broad on substrate range (plastics, composites, primed metals), flexible on flexibility, and medium on chemical/temperature resistance. Silicone rates medium-slow on cure, broad on substrate range including glass and ceramics, very flexible, and high on temperature/UV resistance but low on tear strength [S1][S3][S7][S8]. For a 60-second decision in a fabrication office: rigid structural load, hot/wet service -> epoxy; mixed materials, fast line takt -> MMA; tiny plastic or rubber feet, sub-minute fixture -> cyanoacrylate; expansion-mismatched plastics or primed metal -> PUR; glass, ceramic, or high-temperature gasketing -> silicone.
Common failure modes and process controls
Adhesive-bonded-joint strength is governed by the weakest element in the assembly, not the adhesive itself, so contamination, under-cure, and poor joint design are the dominant causes of in-service failure, not raw adhesive strength [S4]. The most frequently cited shop-floor failures are surface contamination (oils, mold release, fingerprints) under-cure at low ambient temperature, joint designs that load the adhesive in peel rather than shear, and over-potent bondlines that starve the joint of adhesive [S1][S4][S5]. Controls that consistently raise first-pass yield: enforce a documented surface-prep step (solvent wipe plus abrasion where applicable), hold adhesive and substrate above the minimum dispense temperature, design overlap joints to load in shear with a typical overlap-length-to-thickness ratio in the 10-30 range, and use a mix-ratio check (by volume or by weight) for every two-part system [S5][S6]. Specifications should call out fixture time at worst-case ambient, full-cure time, and a coupon-level test method (lap-shear or wedge per ASTM conventions) rather than just a tensile strength number from a datasheet, because datasheet tensile values routinely over-predict real joint capacity.
Standards, sourcing, and qualification
For load-bearing or safety-relevant bonded joints, qualification against recognized test methods (ASTM D1002 lap shear, ASTM D3163 for plastics, ASTM D3762 wedge test, or aerospace/automotive OEM-specific protocols) is a more reliable gate than vendor performance claims [S4][S5]. Chemical and food-contact applications additionally require food-safe epoxy systems with explicit FDA or EU 1935/2004 compliance, and these are listed as a distinct product family by major manufacturers [S3]. Sourcing should track the SDS for region-specific regulatory alignment (Americas, Europe, Asia), and adhesive selection should always cross-reference the SDS set for the specific shop jurisdiction [S3]. The term industrial adhesive used in fabrication catalogs covers a wide range of cure chemistries; matching the chemistry family to the substrate pair and the process window is the engineering work, while the SKU choice is the procurement step. Track two signals over the next sourcing cycle: (a) any shift in OEM guidance toward MMA acrylics for mixed-material vehicle and appliance structures, where fast cure and dissimilar-substrate adhesion are replacing two-part epoxy in select joints, and (b) growth in light-cure acrylates for glass and PMMA display assembly, where 365-405 nm LED curing heads are dropping cycle times below 5 seconds on production lines [S1][S3].
For the relevant spec sheets and selection criteria, see industrial borescope.
Background reading: Polycarbonate Selection for Marine Engineering: Grade, Coating, and Failure-Mode Map.