DSM Somos ProtoGen™ 18420 is a single-component epoxy photopolymer formulated for vat photopolymerization stereolithography systems operating with a solid-state ultraviolet laser at 355 nm. The material is supplied as a filler-free liquid and is intended for layerwise fabrication of functional solid parts requiring moderate elongation at break, high dimensional fidelity, and a fully crosslinked network after secondary ultraviolet postcure. The product designation includes a postcure condition identified as HOC -2; this designation refers to the ultraviolet chamber configuration under which supplier-reported mechanical values are generated, not to a universal dose. Green-state parts removed from the build platform have lower modulus, lower heat deflection temperature, and trapped reactive species. The HOC -2 postcure is therefore part of the product specification rather than an optional processing step. Transfer of the HOC -2 condition to an alternative ultraviolet chamber requires radiometric dose mapping and mechanical property verification coupons because lamp age, reflector geometry, shelf distance, and part orientation alter the delivered surface dose.

Which process variables govern green-state handling before the HOC -2 postcure?

Green-state handling strength is governed by penetration depth D_p and critical exposure E_c. For ProtoGen 18420, supplier processing documentation lists a nominal D_p of 0.14 mm and an E_c near 10 mJ/cm². These values shift with bath age, ambient humidity, and beam uniformity. On production platforms with galvanometer-scanned 250 mW to 500 mW lasers, the working curve should be re-established after every 72 h of idle vat standing because moisture uptake and partial dark advancement can increase apparent E_c. Recoat blade gap settings below 0.10 mm intensify the sensitivity of layer thickness to viscosity drift. A viscosity increase above 350 cP at 30 °C is used as a production control limit because it is an early indicator of moisture contamination or partial cationic advancement. Build platforms with non-uniform irradiance require lasing parameter compensation across the vat plane; failure to compensate produces measurable variation in green-part flexural stiffness before any postcure is applied.

Mechanical property values reported after the HOC -2 postcure are summarized in Table 1. Values are typical, not lot-release minima, and are generated from specimens built in the XY plane at a nominal layer thickness of 0.10 mm. Lot-specific certificate of analysis data should be used for final qualification because batch-to-batch variation in epoxy monomer distribution and photoinitiator content can shift tensile and thermal values.

The heat deflection temperature at 0.46 MPa indicates that ProtoGen 18420 is intended for moderate-temperature service environments. At 1.82 MPa the deflection temperature is lower; published data for this specific datasheet revision is limited, and designers of clamped or bolted assemblies should obtain the full thermal curve rather than extrapolating from the 0.46 MPa value. The notched impact response of this epoxy system is lower than that of high-impact non-epoxy stereolithography grades, and impact performance should not be used as the sole selection criterion for snap-fit or repeated-drop applications.

If a Low-Ash Expendable Pattern Is Required for Investment Casting

Investment casting applications exploit the capability of ProtoGen 18420 to produce expendable patterns with fine surface detail and controlled thermal decomposition during shell burn-out. Burnout schedules should include a controlled ramp through 200 °C to 400 °C because the crosslinked epoxy network undergoes thermal decomposition in that interval. Rapid heating can generate internal pressure ahead of the decomposition front and crack the ceramic shell. Process engineers typically specify a two-stage burnout: a slow ramp at 1 °C/min through 250 °C, followed by a hold at 700 °C to 800 °C for ash removal. Published ash content data for this specific grade is limited; foundries should validate residual ash on a production-scale shell batch rather than relying on laboratory coupon values. Differential thermal expansion between the epoxy pattern and the ceramic shell is managed by filling hollow patterns with a low-density foam or by adding vent passages that prevent pressure accumulation during the early stages of decomposition.

Because stereolithography builds are layerwise, mechanical isotropy cannot be assumed. Tensile specimens built in the XY plane produce the values in Table 1; Z-direction specimens typically show lower elongation at break and lower ultimate tensile strength due to interlayer conversion gradients. On a production line, the Z-direction shortfall is amplified when recoater blade speeds exceed 150 mm/s or when resin temperature falls below 28 °C, both of which increase the probability of incomplete recoat wetting. For parts with snap-fit or threaded features, cross-laminated build orientation and a minimum layer thickness of 0.10 mm are specified to reduce anisotropic failure modes. Prototyping organizations that transfer the same build file from a polyurethane resin to ProtoGen 18420 without re-slicing compensation often report dimensional offsets at thin vertical walls; the offset is traceable to the interaction of the lower green modulus of the epoxy with peel and recoating forces, not to a loss of scan resolution.

Thermal Response and Service Temperature Envelope

Heat deflection under load is the primary thermal limitation for ProtoGen 18420. After the HOC -2 postcure, the network remains in a glassy state at room temperature. Service above the heat deflection temperature leads to creep in clamped joints and loss of dimensional stability. The 0.46 MPa heat deflection temperature of 58 °C is a ranking value and does not define a maximum continuous-use temperature. For dynamically loaded parts, a maximum continuous service temperature below 45 °C is applied when sustained stress exceeds 0.4 MPa; at zero or low stress, brief excursions to 70 °C are typically tolerated without gross distortion, but published data for long-term oxidative stability under these excursions is limited. Postcure under-dosing is the most common cause of thermal underperformance. A part that receives only 80% of the intended UVA dose can exhibit a 5 °C to 10 °C reduction in heat deflection temperature, which is not visually detectable. Ultraviolet radiometers and postcure cycle loggers are required for traceable thermal performance.

Unsealed resin in a production vat absorbs atmospheric water at rates that depend on ambient dew point and air circulation. Water acts as a chain-transfer species in cationic epoxy polymerization and can reduce crosslink density, lower glass transition temperature, and increase surface tack after postcure. Production cells should keep the vat covered during idle periods and maintain the room at 20 °C to 26 °C with relative humidity below 50%. Resin returned from a build platform should be filtered through a 25 µm mesh before re-entering the vat; recycled resin blend ratios above 30% are not recommended without viscosity and working-curve verification. The material should not be exposed to direct sunlight or broad-spectrum ultraviolet room lighting for extended periods, as unintended initiation can produce gel bodies and increase the risk of layer roughness. Solvent exposure should be limited to manufacturer-approved cleaning agents; ketone-based solvents can swell the partially cured network and generate postcure surface microcracking.

Laser Exposure Parameters Do Not Transfer Across SLA Platforms

Direct-write reciprocity between laser scan speed and cure depth does not transfer across stereolithography platforms without beam characterisation. ProtoGen 18420 is processed at 355 nm, but the delivered energy density in the vat plane depends on focused spot diameter, Gaussian beam truncation, scan spacing, and line-to-line overlap. A system calibrated for a 0.10 mm focused spot requires a different exposure per unit area than a system operating at 0.15 mm; simply matching scan speed is insufficient. Production technicians should generate a working curve for each platform and each replacement laser, using the manufacturer-specified D_p and E_c as initial references only. On platforms with variable beam power, daily radiometric verification at the vat surface is required to compensate for laser diode ageing and mirror train contamination. Recoating dynamics for this epoxy are also sensitive to build chamber humidity; a fixed recoating wait time of 3 s per layer is often insufficient at layer thicknesses above 0.15 mm, and experienced operators extend the wait time to 5 s or use a wiper rotation speed below 200 mm/s to eliminate orbital surface defects.

Cleaning of green parts before the HOC -2 postcure is typically performed in isopropanol or a manufacturer-approved solvent blend. Cleaning duration should be limited to 10 min in an ultrasonic bath at 25 °C; longer exposure can plasticize interlayer regions and degrade dimensional precision. After cleaning, parts should be thoroughly dried before ultraviolet postcure to avoid trapping solvent within the network. Entrained solvent can cause blistering or local softening during the exothermic phase of the postcure cycle. Clean parts should be supported in the HOC -2 chamber so that ultraviolet irradiance reaches all functional surfaces; shadowed regions can remain undercured and exhibit reduced heat deflection temperature even when the overall cycle timer has elapsed. Radiometric dose verification should therefore be performed at the lowest-irradiance location within the load, not only at the center of the chamber.

Support structures in stereolithography of this resin require a different contact spacing than in acrylate resins because the green modulus of the epoxy is lower. Supports placed at 3 mm spacing for edge retention and a contact depth of 0.15 mm reduce the occurrence of edge curl on large flat build surfaces. Prime areas below 0.05 mm are more likely to detach from the build platform during peel, particularly on tilt-lift machines with high peel velocities. Dimensional accuracy in unsupported regions benefits from build angles of 10° to 20° relative to the platform, although the final support configuration must be validated for each platform geometry. Residual support nibs after removal should be wet-sanded before postcure; sanding after the HOC -2 cycle is more difficult due to the increased surface hardness of the fully crosslinked epoxy network.

Relative to water-clear stereolithography grades, ProtoGen 18420 is not optimized for transmitted-light optical clarity. Its primary differentiation is the combination of moderate elongation at break and high dimensional fidelity after full postcure. Compared with high-impact non-epoxy grades, this resin shows lower notched impact energy absorption, and published datasheet comparisons indicate that impact values should not be used as the sole selection criterion. Against filled composite stereolithography resins, ProtoGen 18420 avoids filler settling during idle periods but has a lower high-temperature heat deflection temperature; designers who require sustained performance above 60 °C under load should evaluate a different resin grade. For jigs, fixtures, thermoforming masters, and investment casting patterns, the grade is specified where the part is not exposed to continuous ultraviolet radiation, high humidity, or temperatures above the indicated heat deflection threshold. Published data for long-term ultraviolet weathering of this specific formulation is limited; if the part is intended for outdoor deployment, a protective coating or a weathering validation program specific to additive-manufactured epoxy is required. Regulatory conformity statements for REACH and RoHS should be obtained from the supplier’s current certification, because no independent certification is implied by the property table.