Epoxy Resin Technical Support: What We Can Help You Solve Before Ordering

Resin selection for bonded tooling or encapsulated power electronics begins with verification of epoxide equivalent weight and hydrolyzable chloride content rather than generic datasheet acceptance. Standard bisphenol-A diglycidyl ether resins typically exhibit 170–190 g/eq EEW, viscosity of 8–16 Pa·s at 25 °C, and hydrolyzable chloride below 300 ppm according to supplier certificates. A novolac epoxy with an EEW below 180 g/eq raises crosslink density but reduces elongation at break below 2 % when tested per ASTM D638-14; cycloaliphatic resin is specified only when electrical tracking index or UV stability is controlling. For potted assemblies that must survive thermal shock from -40 °C to 130 °C, a flexibilized DGEBA blend with a pre-heat viscosity below 3 Pa·s at 40 °C is reviewed against a coefficient of linear thermal expansion above 60 × 10⁻⁶ K⁻¹ per ASTM E831-19 and wire-bond stress.

Why Does a 5-Weight-Percent Offset in Amine Stoichiometry Create an Under-Cured Interphase?

Alignment of the active hydrogen equivalent weight of the hardener with resin EEW determines the practical mix-ratio window. For an aliphatic amine hardener with AHEW 30 g/eq and a DGEBA resin with EEW 188 g/eq, the stoichiometric ratio is 15.96 phr. A 5 wt% offset lowers crosslink density at the resin–substrate boundary because the catalyst-rich or lean zone changes the rate of oxirane ring opening; after 7 days at 25 °C, the under-cured interphase shows a hardness drop from Shore D 84 to Shore D 76 per ASTM D2240-15. Meter-mix dispensing lines therefore require ratio control within ±2 % by mass, not by volume, because density differences between resin and hardener are typically 0.15–0.25 g/cm³. A production line using a variable-ratio gear pump and a 40:1 L/D static mixer showed that a 3 % ratio drift in a 8,000 mPa·s filled system produced a glass transition temperature drop of 6 °C after cure, measured per ASTM D3418-21. Published data for this specific configuration is limited to this measured shift; the broader implication is that static mixers with 24–32 elements cannot compensate for poor ratio control when the hardener is fast at 25 °C.

Temperature-dependent viscosity, not the datasheet value at 25 °C, determines whether a 0.5 mm gap can be impregnated without dry-fiber regions in carbon fiber laminates. A filled DGEBA system with a room-temperature viscosity of 12,000 mPa·s can fall below 600 mPa·s at 50 °C, but heating reduces pot life from 45 min to 12 min in a 1 kg mass. Vacuum deaeration at 1–5 mbar is specified only after resin and hardener are pre-mixed for 90–120 s at 800 rpm in a planetary mixer; higher shear entrains air and extends the time to clear film below 1 mm. For filled systems, pre-order support examines the thixotropic index from a Brookfield viscometer sweep at 2 rpm and 20 rpm per ASTM D2196-20, because a thixotropic index below 2.0 allows filler settling in a 1 kg pail before dispensing.

Mass Casting Thickness, Mold Thermal Conductivity, and the Nonlinear Runaway Boundary

In mass casting, the critical thickness for thermal runaway is not a fixed property; it depends on mold thermal conductivity, initial resin temperature, and hardener type. For unfilled DGEBA/aliphatic amine at 30 °C, a 25 mm section in a steel mold can exceed 180 °C at the core, while the same section in an aluminum mold may peak 30–40 °C lower because of higher heat dissipation. Production experience on a 15 kg potting job for a medium-voltage transformer showed that casting in 5 kg lifts at 30-minute intervals kept internal temperature below 110 °C while maintaining a Shore D 82 hardness after 24 h at 25 °C. Published data for exact peak temperatures in all mold configurations is limited; oven post-cure should not begin until the part surface temperature drops below 60 °C to avoid boundary cracking. Pre-order support therefore specifies the maximum exotherm temperature limit as a function of section thickness, not as a single gel time.

With aluminum trihydrate at 100 phr, mixed viscosity increases from 1,800 mPa·s to 14,000 mPa·s at 25 °C. Centrifugal deaeration or three-roll milling is used before fumed silica is added at 2–3 wt% for anti-settling control; a cone-plate rheometer sweep from 1–100 s⁻¹ is reviewed to determine whether a 10 μm aluminum oxide filler will settle 0.5 mm after 2 h. Compounding a ceramic-filled encapsulant on a co-rotating twin-screw extruder with L/D 40:1 and barrel zones at 60/70/80/80/80/75 °C reduces filler agglomerates below 5 μm, but increases moisture pickup; pre-drying at 60 °C for 4 h is required when ambient relative humidity exceeds 60 %. Without this data, fillers settle before gelation and produce a hardness gradient across the part.

Post-Cure Window Sensitivity in Anhydride Systems: DSC Enthalpy and Residual Monomer

Anhydride-cured DGEBA systems do not reach full conversion without a post-cure. Differential scanning calorimetry per ASTM D3418-21 after 24 h at 25 °C typically leaves 10–15 % residual oxirane when a latent accelerator is absent; post-curing 4 h at 120 °C increased glass transition temperature from 62 °C to 118 °C in a methylhexahydrophthalic anhydride system. Production lines with conveyorized infrared ovens must hold the part above 110 °C for at least 2 h because diffusion of unreacted oxirane falls rapidly below this threshold. Premature exposure to hot oil at 130 °C before post-cure creates a crosslinked skin that traps unreacted hardener; salt-spray resistance after 1,000 h per ISO 9227:2022 then drops below the 2 mm maximum creep from the scribe line.

Adhesion to cold-rolled steel with a roughness of Rz 20 μm is not governed by resin chemistry alone; the surface must be grit-blasted to ISO 8501-1 Sa 2½ and degreased before primer application. Lap shear specimens prepared per ASTM D1002-10 from 1.6 mm aluminum alloy show tensile shear strength above 18 MPa only when a silane adhesion promoter is used; without the promoter, failure shifts from cohesive to adhesive at 50–70 % relative humidity. On a production line bonding polycarbonate enclosures, a flame treatment raising surface energy from 34 mN/m to 46 mN/m increased pull-off adhesion per ASTM D4541-17 from 1.5 MPa to 8 MPa. Amine blush must also be excluded before secondary bonding: at 22 °C and 70 % RH, a cycloaliphatic amine hardener produces surface carbamate salts within 60 min, detectable by a water-break test and requiring mechanical removal before adhesive application.

For Acidic Media, Diffusion-Limited Mass Uptake and the Threshold for Property Collapse

Prolonged immersion in 10 % sulfuric acid at 60 °C yields mass uptake below 1.5 % after 28 days for an aromatic amine-cured novolac system, while a standard DGEBA/aliphatic amine control exceeds 4 % mass gain under the same exposure and loses 22 % of its original flexural strength when tested per ASTM D790-17. Acetone exposure is more aggressive: 7-day immersion at 25 °C causes visible softening in DGEBA systems with aliphatic amine hardeners, but an anhydride-novolac system retains 85 % of its flexural modulus. The technical review selects the chemical exposure table only after the exact solvent concentration and temperature are known; a generic “chemical-resistant epoxy” statement is not valid without diffusion rate data.

Electrical insulation specifications before ordering require simultaneous review of volume resistivity and comparative tracking index because a high dielectric strength value does not prevent surface tracking under humid contamination. A cycloaliphatic anhydride system may show a dielectric strength of 18 kV/mm per ASTM D149-20 and a comparative tracking index of 600 V per IEC 60112:2020, whereas a brominated DGEBA system with 8 % antimony trioxide drops below 400 V CTI under the same test. For power modules operating at 1.2 kV, partial discharge inception voltage is evaluated on cast samples aged at 85 °C/85 % RH for 1,000 h; systems with glass transition temperatures below 120 °C show a shift in dielectric loss because absorbed water couples to the network at high frequency.

Linear cure shrinkage of a standard DGEBA amine system is between 0.8 % and 1.5 % by volume depending on filler load; this creates internal stress in potted assemblies when the potting thickness exceeds 10 mm. A filled system with 65 wt% fused silica lowers shrinkage to below 0.4 % and reduces coefficient of linear thermal expansion to 25 × 10⁻⁶ K⁻¹ per ASTM E831-19. For epoxy molding compounds in transfer molding tools with 50 t clamp force, low-shrinkage novolac systems reduce wire sweep and die-shift failures because the melt-front stress magnitude is lower at gate velocities below 20 mm/s.

When Optical Clarity Requirements Collide with Aromatic Hardener Yellowing

Transparent encapsulation requires a low-color cycloaliphatic resin combined with a non-aromatic anhydride hardener, because aromatic amine and bisphenol-A backbones absorb in the 350–400 nm range. After 1,000 h of UVA-340 exposure per ASTM G154-23, a cycloaliphatic anhydride system may retain 92 % of its initial transmittance at 450 nm when a hindered amine light stabilizer is included at 0.5 wt%. The same formulation with 0.1 wt% antioxidant shows a yellowing index increase of 6 units per ASTM E313-20. Technical review also covers cure schedule: rapid curing at 120 °C without a gradual ramp can create microvoids from hardener volatility, producing haze that no subsequent optical coating can correct.

Flame retardance is evaluated only at the finished-part thickness because UL 94 V-0 at 3.0 mm does not transfer to a 0.8 mm wall section. A brominated epoxy with 20 % bromine content passes UL 94 V-0 at 3.0 mm when antimony trioxide is used at 4–6 wt% as a synergist; phosphorus-based alternatives require 2–3 wt% phosphorus and may reduce dielectric strength by 10–15 %. Pre-order support includes a thermal analysis of the cured compound by ASTM E1354-22 cone calorimetry only when the end product is installed in an enclosed plenum or rail vehicle interior.

Regulatory compatibility for potable-water contact is verified by reviewing cured formulations against NSF/ANSI/CAN 61 for extraction of bisphenol A and epichlorohydrin, not by assuming that all food-contact epoxies are equivalent. An epoxy novolac used in a solvent storage lining may meet FDA 21 CFR 175.300 for dry food contact, but the same unpigmented system can fail European food-contact migration limits when tested under 10 % ethanol simulant at 40 °C for 10 days. For electrical assemblies shipped into the EU, the formulation is checked against the REACH candidate list and RoHS Directive 2015/863/EU for restricted phthalates, lead, and cadmium; a filled system containing 5 wt% carbon black may still comply with RoHS if the carbon black itself is free of polycyclic aromatic hydrocarbons above the supplier’s threshold. Compliance support before ordering includes requesting full disclosure of the curing agent’s amine value and free monomer content, not only the safety data sheet.