Epoxy resin grades E-51, E-44, and E-20 are compared through three acceptance criteria: epoxy equivalent weight (EEW), viscosity, and softening point. All three grades are bisphenol A diglycidyl ether resins classified under GB/T 13657-2011, with the grade numeral representing the nominal epoxy value multiplied by 100. E-51 is a low-molecular-weight liquid with an EEW range of 185–210 g/eq and a rotational viscosity of 10000–16000 mPa·s at 25 °C. E-44 is a semi-solid intermediate with an EEW range of 210–240 g/eq and a ring-and-ball softening point of 12–20 °C. E-20 is a solid flake or chip with an EEW range of 455–556 g/eq and a softening point of 64–76 °C. These acceptance ranges control pumping, heating, dissolution, and curing-agent demand in downstream manufacturing.

What distinguishes viscosity from softening point?

Viscosity is measured on E-51 by ISO 2555:2018 using a single-cylinder rotational viscometer at 25 °C; softening point is measured by ISO 4625-1:2020 ring-and-ball method. The shift from viscosity to softening point is not a simple test-preference change. It reflects the molecular weight at which ambient thermal energy no longer allows gross flow. E-44 at 15 °C may present as a tacky semi-solid, while at 40 °C it becomes pumpable; E-20 remains solid until approximately 64 °C and requires melt work above that point. A viscosity datum alone would misrepresent E-20 because the temperature needed for flow depends on molecular weight distribution, not just average EEW. Rotational viscometer readings must be considered in terms of shear rate: E-51 is nearly Newtonian in the 10–160 s⁻¹ window, whereas filled E-44 compounds are shear-thinning. Softening point repeatability within a single laboratory is typically better than ±0.5 °C when the same ring, ball, and heating medium batch are used. In batch control, softening point is more sensitive to oligomer chain growth than EEW, especially when resin has been overheated and advanced during storage.

Comparative acceptance values for the three grades are shown below. The EEW ranges follow the product standard, and the viscosity or softening point column identifies which physical form dominates the specification. Individual supplier certificates of analysis may fall within narrower quality-control bands, but the standard ranges remain the purchasing boundary.

Crosslink density and backbone extent as functions of EEW

Under idealized stoichiometric cure, reactive group concentration scales inversely with EEW. The DGEBA repeat unit adds approximately 142 g/eq per bisphenol A segment. E-51 at nominal EEW 196 g/eq has an oligomer extension n of approximately 0.18; E-44 at 227 g/eq has n≈0.40; E-20 at 500 g/eq has n≈2.32. The calculated oxirane content falls from 5.1 mol/kg for E-51 to 4.4 mol/kg for E-44 and 2.0 mol/kg for E-20. Stoichiometric hardener demand is obtained as AHEW/EEW × 100. For a hardener with AHEW 100 g/eq, E-51 requires 51 phr, E-44 requires 44 phr, and E-20 requires 20 phr. A ±2 phr weighing error changes the epoxide-to-amine ratio by approximately 4% relative to E-51, which is sufficient to alter gel time and final glass transition temperature when measured by ISO 11357-2:2020. The exact initial EEW should be rechecked by titration per ISO 3001:1999 before each large batch because stored E-51 can advance by 1–5 g/eq over 6 months at 25 °C in partially opened containers.

Adiabatic temperature rise in unpigmented E-51 remains a processing limit. Because E-51 delivers the highest oxirane concentration of the three grades, a 100 g mass cured with a fast cycloaliphatic amine can exceed 180 °C in a poorly cooled mold. Industrial casters therefore restrict unfilled E-51 pours to cross-sections below 25 mm and load fused silica at 50–65 wt% to reduce exotherm and shrinkage. Thermocouple data from a 50 mm cylindrical mold are required because centerline-to-wall gradients above 20 °C produce non-uniform gelation and internal stress. E-20, with its lower oxirane molarity, does not generate the same adiabatic peak; however, the higher melt viscosity imposes its own mixing limitation. Published heat-rise data for every hardener combination is limited, so thermal profiling in the specific mold geometry is required before scaling.

When E-20 requires premelting, screw torque responds nonlinearly

On production-scale powder coating premix lines, E-20 flake is metered into a 70 mm co-rotating twin-screw extruder with L/D 40:1 and barrel zones set at 90–120 °C. If the flake enters the feed throat below its softening point, the first kneading block receives solid particles rather than a compressible melt. Motor torque rises because the screw must crush and convey solid cylindrical blocks before viscous dissipation completes melting. The response is nonlinear; a feed-throat temperature drop from 60 °C to 50 °C may produce a larger torque increase than a drop from 90 °C to 80 °C because the material crosses the softening point. The resulting powder coating after milling can show pigment agglomerates, poor electrostatic charge retention, and gel defects. Published data for this specific screw configuration is limited; the operational pattern is described in supplier processing bulletins and avoided by preheating E-20 to 70–80 °C before extrusion.

Production batches of E-44 are stored in stainless steel vessels with low-shear anchor agitation at 20–40 rpm and jacket temperature 35–40 °C. If jacket temperature falls below 20 °C, the resin near the wall can solidify and the anchor can rotate in a channel, reducing heat transfer and leaving the core at 25 °C for extended periods. Drum thawing at 50–60 °C requires 18–24 h for a 200 L drum because heat must penetrate from the surface inward. Transfer lines below 25 mm nominal bore are not advisable for E-44 under cold conditions unless steam-jacketed or electrically traced at 40 °C. In filled casting compounds, E-44 is often selected because the semi-solid plateau reduces filler sedimentation during batch holding; however, the higher initial viscosity requires vacuum mixing at 40–50 °C and 1–5 kPa absolute pressure for 30 min to achieve bubble release.

Solvent cut is not a cold-cut operation for E-20

The cut of E-20 into solvent is performed at 80–100 °C under reflux; cold cutting below the softening point is not effective because the solid flake remains dispersed as gel bodies. E-51 is commonly reduced with reactive diluents or aromatic solvents at ambient conditions, but E-44 requires warm solvent at 50–60 °C and may crystallize if diluted below 70% solids and stored below 10 °C. For E-20 high-solids systems, aromatic hydrocarbon content below approximately 60% of the solvent blend can precipitate the resin and clog 100–250 µm filter elements. Moisture ingress in solid E-20 flake stored at relative humidity above 60% can introduce 0.05–0.15% water, which is sufficient to hydrolyze dicyandiamide during extrusion and produce cratering in powder coatings. Predrying is therefore required at 40–50 °C for 2–4 h or in a vacuum dryer at 70 °C and 20 kPa. E-51 and E-44 must be segregated from amine hardeners and strong tertiary amine accelerators during storage because premature chain extension raises viscosity and reduces epoxy value.

Wet lay-up lamination with E-51 uses vacuum degassing at 0.5–1 kPa for 10–15 min before fabric impregnation. The mixed system remains below 500 mPa·s for 20–30 min at 25 °C, allowing saturation of 200 g/m² unidirectional carbon fabric without void formation. In high-solids epoxy ester can coatings, E-20 is dissolved at 80–100 °C and panels are sprayed at 8–12 µm dry film thickness; solvent resistance is evaluated by ASTM D5402-19 methyl ethyl ketone double-rub resistance. For E-20 powder coating extrusion, line startup is not permitted until the resin hopper and feed throat have reached 70 °C.