DSM Somos ProtoGen™ 18920 is a liquid photopolymer supplied for 355 nm stereolithography systems. The material enters the process as a low-viscosity, opaque white reactive liquid and is converted through a two-stage cure: first by laser-initiated crosslinking in the vat, then by the designated HOC +3 ultraviolet postcure. The full trade descriptor, “DSM Somos ProtoGen™ 18920 Liquid Photopolymer, UV Postcure at HOC +3,” is a process instruction rather than an optional finish. It identifies a grade that requires a specific high-optical-output UV cycle after green-state cleaning and drying.

Because the material is a thermoset-forming liquid, it is not processed through a twin-screw extruder or an injection molding machine. Production-equipment behavior is defined by the recoater blade, the 355 nm laser scan strategy, and the UV flood chamber. Incoming inspection typically records viscosity at 30 °C and density at 25 °C. Typical values for this product family are approximately 250 mPa·s and 1.13 g/cm³, respectively; these values are not batch guarantees and should be verified against the certificate of analysis.

Before use, the resin is stirred gently for 15 min and filtered through a 100 µm mesh into the vat. Storage is maintained at 15 °C to 30 °C in the original opaque container. Ambient or actinic light exposure generates gel bodies that can form surface defects or jam the recoater blade. On high-throughput lines, viscosity drift above 350 mPa·s at 30 °C has been associated with recoater chatter and horizontal banding on the part surface.

Viscosity is normally checked with a rotational viscometer under ISO 2884-1:2016 or ASTM D2196-20, while density is determined under ISO 1183-1:2019 or ASTM D792-20. These measurements are used for incoming inspection, not for part design.

The grade is specified for opaque ABS-like prototypes, vacuum-casting masters, wind-tunnel test models, jigs, and patient-specific medical device models. The material is selected where optical clarity is not required and where downstream operations include drilling, tapping, bonding, or painting. In fluid-flow manifold prototypes, the opaque surface can improve structured-light scanning contrast compared with translucent grades, although the surface must be covered with a matting spray if gloss reflections exceed the scanner’s dynamic range.

Mechanical property comparisons are performed on postcured bars conditioned under ISO 291:2008 class 2 atmosphere and tested under ISO 527-2:2012 or ASTM D638-14 for tensile response, ISO 178:2019 or ASTM D790-17 for flexural response, and ISO 75-2:2013 or ASTM D648-18 for heat deflection temperature. Izod impact is method-specific; ASTM D256-10 values cannot be converted directly to ISO 179-1:2023 because the specimen cross-section and notch tip radius differ.

What Distinguishes ProtoGen 18920 from Clear WaterShed XC 11122 and High-Stiffness Ceramic-Filled Resins?

Direct substitution of published ProtoGen 18120 data for ProtoGen 18920 is not permitted for qualification because the 18920 formulation can differ in photoinitiator package, stabilizer content, and postcure response. Where preliminary design permits approximation, the user should treat the data as class-level rather than grade-level. Published data for this specific configuration is limited outside the manufacturer’s technical datasheet; lot-specific testing closes the gap.

The difference from clear grades such as WaterShed XC 11122 is primarily optical and process-related. Clear resins allow internal flow visualization but are more sensitive to clouding from solvent retention and moisture. Opaque white grades scatter light, which shortens optical penetration depth and may require a higher laser exposure or reduced scan speed for a given layer thickness. The practical outcome is that ProtoGen 18920 is often easier to inspect by surface metrology and can conceal internal lattice support witness marks, but it cannot be used for transparent fluid-flow models.

Compared with high-stiffness ceramic-filled materials such as Somos PerFORM, ProtoGen 18920 has a lower modulus and lower heat deflection temperature; it is not a substitute for composite cure tooling that must survive autoclave or press cycles above 120 °C. Its lower filled content also makes it less abrasive to machining tools and less likely to settle in the vat, but the trade-off is lower creep resistance under sustained load at elevated temperature.

Postcure Dose Response and Dimensional Stability Across Wall Thickness Ranges

The HOC +3 ultraviolet postcure is not a gross overexposure step. It is a defined irradiation and dwell schedule intended to raise residual conversion above the threshold where creep and solvent sensitivity become negligible. In air-atmosphere chambers, the acrylate component remains susceptible to oxygen inhibition at the part surface. If the chamber irradiance decays by more than 20 % from the calibrated value, the outside 50 µm to 100 µm can remain under-converted even when the core is cured. Calibration with a radiometer in the 320 nm to 390 nm band is therefore part of lot release for production parts.

Residual conversion can be monitored by differential scanning calorimetry in nitrogen at a ramp rate of 10 °C/min. An undercured part typically shows an exotherm associated with unreacted acrylate or epoxy groups. The test is useful when the HOC +3 protocol is transferred across machines because chamber geometry, lamp age, and part packing density affect the actual dose received. A part stacked in a shadowed zone may receive a dose below the chamber’s calibration setpoint and can exhibit anisotropic conversion from top to bottom.

Thermal distortion during postcure arises from differential conversion between the outer skin and the core. A thin wall below 1 mm can reach full conversion early and become brittle, while a thick boss remains under-converted and generates shrinkage stress at the interface. Supporting critical walls with ribs, reducing the layer thickness to 50 µm, and allowing the green part to rest at 23 °C for 1 h before postcure reduce the tendency for warpage. Dimensional metrology should be performed after the part has cooled to 23 °C and been conditioned at 50 % RH for 24 h; moisture absorption can shift dimensions by several hundred micrometres per metre in polyamide-like photopolymers, and the exact drift should be measured for the lot.

Green-state cleaning uses isopropanol or the higher-flash-point solvent specified in the manufacturer work instruction. Extended immersion swells the partially cured network and can lead to surface crazing after the HOC +3 cycle. On automated cleaning lines, sequential soak and spray cycles are preferred over long solvent baths because they remove uncured monomer from blind pockets without saturating thin sections.

In design-for-SLA analysis, the process window for ProtoGen 18920 is constrained by the laser exposure required to retain the trace and the thermal stability required to survive the HOC +3 cycle. A 50 µm layer is recommended when the part contains snap-fit arms, living hinges, or push-fit bosses because the lower layer height reduces stair-step error and creates a smoother surface before postcure. A 100 µm layer reduces build time but increases the amount of uncured material trapped between layers; consequently, draft builds at 100 µm are usually limited to parts without closed internal channels.

The build platform should be leveled according to the SLA machine manual. Resin vat temperature during build is held between 28 °C and 30 °C. If the temperature falls below 20 °C, viscosity increases and the recoater blade can float, producing layer thickness error. Infrared preheaters or vat heaters are preferred over direct resin preheating because local overheating can advance polymerization and change the exposure response.

After the build, the green part is drained over the vat for 10 min to 20 min to reduce solvent loading. Supports are removed before postcure because the differential shrinkage between the support lattice and the part can initiate contact-point fissures. For thin walls below 1 mm, support tips are reduced and support density is increased to prevent peel from the build platform during recoating.

On 355 nm platforms such as 3D Systems SLA 3500, Viper Si2, and ProX 800, the material can be used with standard recoating and scanning parameters after an initial exposure adjustment. The switch from a clear resin to an opaque white grade typically reduces the optical penetration depth; therefore the scan speed or laser power may need to be adjusted to keep the same cure depth. Operators should not copy a clear-resin parameter file without a test build because the resulting green part can be undercured at the base layers and may delaminate from the platform.

When the HOC +3 Protocol Is Applied Before Medical Device Biocompatibility Testing

The HOC +3 step is a material processing operation, not a sterilization or cleaning step. For medical devices, the finished part must be evaluated under ISO 10993-1:2018 with endpoints selected from the device duration and contact type. Cytotoxicity testing under ISO 10993-5 is typically the first screen because reactive monomers and photoinitiator fragments can leach from undercured surfaces. Extract preparation should follow ISO 10993-12, and the postcure history must match the production process exactly; a change in UV dose or chamber loading can alter the leachable profile even if the part appearance is unchanged.

For skin-contacting devices, irritation and sensitization data under ISO 10993-10 are required, along with lot-specific documentation from the resin supplier. The liquid monomer is not biocompatible by virtue of the grade name, and the safety data sheet should be consulted before handling. If steam sterilization is specified, the effect of elevated temperature on dimensional stability must be tested because the heat deflection temperature of the postcured resin places the material in a low-temperature service class; ethylene oxide or hydrogen peroxide gas plasma may be considered only after material compatibility testing.

From a regulatory standpoint, the resin itself is not an electrical/electronic equipment; therefore RoHS Directive 2011/65/EU applies to the finished device rather than the liquid. Supplier declarations should cover lead, mercury, cadmium, hexavalent chromium, PBB, and PBDE for the cured polymer after postcure. REACH obligations are addressed through safety data sheet Section 15 and are lot-dependent; no blanket statement should be made without review of the latest safety data sheet.

The material is generally not suited for prolonged outdoor exposure without a paint or clear coat because the remaining network can further crosslink under sunlight and become brittle. For outdoor applications, UV stabilization should be qualified under ISO 4892-2:2013 or ASTM G154-16, and the test should compare postcured and painted specimens. Indoor environments with temperatures below 40 °C are less likely to exceed the mechanical service window of the resin.

Solvent compatibility after postcure is limited for strong polar solvents and aromatic hydrocarbons. Immersion testing should follow ISO 175:2010 for the intended service fluid because the partially polar network can absorb solvents and soften. Isopropanol, ethanol, and aliphatic hydrocarbons are often used for wipe-down operations, but ketones and chlorinated solvents should be avoided unless validated.

Machining of postcured ProtoGen 18920 is performed with high-speed steel or carbide tooling at low feed rates and low cutting temperatures. Since the material is thermoset, it does not melt like ABS; it can chip or fracture if the tool is dull or the cut is too aggressive. Cutting parameters should be developed experimentally because no single ISO test captures machinability across all solid-slab and thin-wall geometries.

On production lines, three failure modes are commonly observed when the material is transitioned from prototype work to low-volume manufacturing: residue entrapment in blind channels, postcure distortion of unsupported thin walls, and batch-to-batch viscosity drift. The first is controlled by adding drain holes or rotating the part during cleaning. The second is controlled by support design and by limiting the unsupported span before postcure. The third is controlled by incoming viscosity measurement and by mixing settled material before the build.

The product should not be mixed with other photopolymer grades unless the manufacturer has validated the mixture. Mixing with dissimilar resins can produce hazing, soft zones, or gel bodies that adhere to the recoater blade. If a layer is skipped due to a floating recoater, the operator should stop the build and inspect the blade gap; continuing without correction can produce a horizontal delamination that is not recovered by postcure.

Published open-literature data for the exact HOC +3 radiant dose and dwell schedule are limited; the protocol must be taken from the manufacturer’s process sheet or machine-specific work instruction linked to the lot. Operators should treat HOC +3 as a controlled process variable and record chamber temperature, lamp power, and cycle duration in the batch record.