Commercial epoxy resin adhesives are most often formulated from liquid diglycidyl ether of bisphenol A (DGEBA) with an epoxide equivalent weight of 170–190 g/eq. The cured mechanical and adhesion properties are governed primarily by three process variables: mix ratio, cure schedule, and substrate preparation. The following guidance functions as an Epoxy Adhesives: Mix Ratio, Cure Schedule & Substrate Prep Cheat Sheet for laboratory and production bonding operations. In a two-part system, the correct mass ratio is derived from the hardener equivalent weight rather than from any fixed 2:1 or 1:1 convention. Amine-cured systems follow stoichiometry between epoxide groups and amine hydrogen atoms, and the mixed adhesive should be verified by gel time per ASTM D2471 and mixed viscosity per ASTM D2196 at 25 °C before bonding. Standard DGEBA resins are limited to continuous service below 60–80 °C unless post-cured with cycloaliphatic amine or anhydride chemistry. Epoxy novolac resins are used where glass transition temperatures above 150 °C and broad chemical resistance are required; their average functionality of 2.2–2.7 epoxide groups per molecule raises crosslink density but also increases mixed viscosity and requires higher mix ratios.

Hardener class dictates pot life, cure temperature, and moisture sensitivity. Low-molecular-weight aliphatic amines such as triethylenetetramine provide fast room-temperature cure but can cause skin sensitisation and high exotherm. Polyamidoamine hardeners provide lower modulus and good adhesion to damp surfaces but cure slowly and develop glass transition temperatures in the 40–60 °C range. Cycloaliphatic amines such as isophoronediamine give glass transitions of 90–120 °C after post-cure and lower moisture absorption than aliphatic counterparts. Anhydride hardeners require schedules such as 2 h at 100 °C plus 2 h at 150 °C and are selected for high electrical insulation and very low mixed viscosity. Each class therefore demands a separate mixing ratio, not a standard approximation.

How Do Off-Ratio Dispensing Errors Alter the Glass Transition and Interfacial Strength?

The stoichiometric weight ratio for 100 g of DGEBA with an epoxide equivalent weight of 190 g/eq and a hardener with a hydrogen equivalent weight of 24 g/eq is 12.6 phr. If the same resin batch shows an EEW of 170 g/eq, the ratio shifts to 14.1 phr; this 1.5 phr shift is sufficient to alter gel time and final crosslink density. Dispensing should therefore be based on the certificate of analysis for each batch. A mass error of ±2 wt% is often tolerable for noncritical overlays, but off-ratio conditions above ±5 wt% produce distinct failure modes. Excess amine hardener remains available at the surface and forms carbamate compounds in humid air, creating a waxy or greasy interphase that can reduce pull-off adhesion in ASTM D4541-22 by more than 30%. Excess epoxy leaves unreacted epoxide groups and chain ends; differential scanning calorimetry per ISO 11357-2:2020 shows a glass transition reduction of 5–15 °C when the hardener is undercharged by 10%. Because resin and hardener densities can differ by 10–20%, volume-based mix cups amplify ratio drift; mass-based dispensing with an accuracy of ±0.1 g is required for component masses below 50 g.

These representative values must be recalculated when active diluents, tougheners, or thixotropes are added. A monofunctional reactive diluent at 10 phr lowers the effective EEW and modifies crosslink density; each formulation change therefore requires a new stoichiometric calculation and a gel time check rather than a simple lookup.

Substrate preparation below the adhesive line is the most common source of field failure. On low-carbon steel, solvent cleaning per SSPC-SP1 alone leaves mill scale, oils, and weak oxide layers. Near-white metal blast cleaning per SSPC-SP10/NACE No. 2 followed by an angular-profile of 75–100 µm verified to ASTM D4417 provides mechanical interlock for high-modulus epoxy adhesives. Soluble salt contamination should be below 20 mg/m² as chloride, measured by ISO 8502-6; higher values drive osmotic blistering at the interface. The cleaned steel should be bonded or primed within 4 h at 25 °C and 60% RH before visible re-oxidation occurs. Aluminium bonding requires either abrasive blasting or chemical etch per ASTM D2651, followed by primer within 4 h; water-break-free inspection should be used to verify that the surface energy is adequate. Concrete must be free of laitance and oil. Surface profile CSP 3–5 under ICRI 310.2R-2020 is typical, internal relative humidity must remain below 80% per ASTM F2170, and surface pH should be 10 or lower. Thermoplastic substrates such as polypropylene are not bondable by solvent wipe alone; flame or corona treatment to a surface energy of 38–48 mN/m per ASTM D2578 is required. The treated surface must be bonded within 30 min because mobile low-molecular-weight species migrate back to the surface and reduce wetting.

On concrete, the substrate itself can fail before the epoxy. A pull-off strength of 1.5 MPa or higher on prepared concrete per ASTM C1583 is a common acceptance level for structural epoxy bonding; lower values indicate the surface must be mechanically removed and re-profiled rather than primed. On metal, the blast profile should not exceed 100 µm for thin bondlines because excessive profile can prevent close contact and create voids at the interface.

When Bondline Temperature Falls Below 10 °C, the Cure Schedule Must Be Extended by 4–6× or a Post-Cure Must Be Added

Cure schedule cannot be selected only from ambient air temperature; the actual bondline temperature controls reaction rate. For aliphatic amine-DGEBA systems, the gel time at 25 °C is 20–40 min, but at 10 °C gel time may exceed 90 min and full property development may require 7–14 days. Below 5 °C, polyamidoamine systems often require 14–21 days to reach film cure, and the interphase can plasticise when exposed to moisture. Catalytic or low-temperature hardeners can extend the practical lower limit to approximately 0 °C, but published data for specific formulations at sub-zero application is limited. Higher temperature accelerates cure, but bondlines above 3 mm can autotherm; the adiabatic temperature rise of neat DGEBA-amine mixtures can exceed 100 °C, producing voids and amine blush. The cure acceptance criterion for structural bonding is usually a residual exotherm of ≤5 J/g by DSC per ISO 11357-2:2020 and a Tg within 10 °C of the fully cured system. Post-cure of 60–100 °C for 1–2 h after ambient gel is common for cycloaliphatic and anhydride systems but should not be applied to already vitrified joints without first heating the substrate.

The vitrification limitation can be understood from time-temperature-transformation behaviour. At 25 °C, the glass transition of the curing network eventually reaches the cure temperature, molecular mobility becomes diffusion-limited, and the reaction slows even though epoxide and amine groups remain. That is why room-temperature-cured DGEBA systems often plateau at 80–95% conversion. Post-cure above the ultimate Tg restores mobility and drives residual exotherm below 5 J/g. Apparent activation energy for a typical aliphatic amine-DGEBA reaction is in the range of 45–60 kJ/mol; the practical effect is that cure time doubles or halves per 10 °C change within the 10–40 °C interval. Adhesive batches cured below 15 °C should therefore be scheduled for either an extended room-temperature dwell or an oven ramp, not released on the basis of surface tack alone.

Pot life is not identical to gel time. In a 100 g mixed mass, gel time by ASTM D2471 may be 30 min, but hand application becomes difficult after 15–20 min because viscosity doubles from an initial 2 Pa·s to 4–6 Pa·s. In thicker sections, the exotherm reduces viscosity before crosslinking, which can extend the apparent open time while shortening gel time. For static-mix dispensing, the mixed viscosity should remain below the system limit of the metering pump; many production systems are limited to 20 Pa·s at the mix outlet. Filled epoxies with silica, aluminium oxide, or glass bubbles exhibit shear-thinning behaviour. Viscosity should be measured at the relevant shear rate, typically 20–100 s⁻¹, rather than at low-shear Brookfield values. Batch-to-batch hardener viscosity should be checked per ASTM D2196 at 25 °C; a shift greater than 15% indicates moisture uptake, carbon dioxide absorption, or hardener ageing.

Viscosity Rise, Filler Settling, and Ratio Drift in Meter-Mix Production

Production meter-mix equipment introduces additional failure modes that are not observed in cup mixing. Static mixers with 24–32 elements and outlet pressure of 10–20 bar homogenise filled epoxy pastes, but pressure fluctuation above ±0.5 bar across a 500 mL cartridge can shift the resin-hardener ratio by 3–5% by weight. Hardener reservoirs exposed to humid air absorb carbon dioxide and water, forming carbamate and increasing viscosity; batch records should flag a viscosity shift greater than 15% at 25 °C per ASTM D2196. Check-valve wear in positive-displacement pumps produces intermittent ratio spikes that are difficult to detect from visual bead appearance alone. In manual production quantities, a two-container mass check with an accuracy of ±0.1 g is necessary for component masses below 50 g because small batch error scales directly into stoichiometric imbalance. Bondline thickness for high-modulus structural epoxies should be maintained at 0.1–0.5 mm; thicker sections reduce lap shear strength and increase void populations, as measured by ASTM D1002 and ISO 4587.

Moisture above 85% RH during application can create amine blush on aliphatic and cycloaliphatic amine systems even when the mix ratio is correct. The blush can be removed from cured excess hardener by water washing followed by light abrasion, but in production it is easier to switch to a blush-resistant hardener or lower ambient humidity. Avoid direct combination of amine-cured epoxies with uncured polyurethane residues or ester-based release agents because interfacial migration can compete with epoxy crosslinking and reduce adhesion. On galvanised steel, surface preparation must remove the zinc oxide layer and apply a compatible primer; direct epoxies can react with zinc corrosion products under humid conditions. These boundaries are not formulation limitations alone; they affect the same mix ratio, cure schedule, and substrate preparation checklist that governs the final adhesive bond.