Industrial stereolithography resins formulated for water contact and moderate thermal exposure occupy a narrow processing band between general-purpose ABS-like photopolymers and ceramic-filled high-temperature systems. DSM Somos ProtoTherm™ 12120 is supplied as a liquid photopolymer for vat photopolymerization on 355 nm solid-state laser platforms, with the thermal postcure step specified as mandatory. In the green state the material behaves as a rigid thermoset with high green strength and limited resistance to polar fluids; after a prescribed thermal postcure cycle the crosslink density advances, water absorption stabilizes, and the heat deflection temperature shifts upward. Manufacturer-published technical literature reports a postcure heat deflection temperature at 0.46 MPa close to 121°C, consistent with the 12120 model designation. The resin is used for low-volume functional parts exposed to humid air, water spray, and intermittent immersion where dimensional drift in standard epoxy-acrylate SL materials causes tolerance failure.

What Reduces Green-State Dimensional Variability in Thick Cross-Sections?

Large-section parts built with ProtoTherm 12120 accumulate shrinkage stress during layerwise photopolymerization. On 3D Systems iPro 8000 vat platforms, technicians commonly reduce solid cross-section thickness or apply hollow-and-drain structures to limit delamination at build start. A build temperature between 28°C and 32°C is maintained to keep viscosity low enough for blade recoating, while laser power compensation is adjusted for beam diameter changes across the platform. Supports for walls thicker than 4 mm are placed at 0.5 mm to 1.0 mm spacing along planned machined surfaces, because dense support spacing increases force during part removal and can introduce microcracking at the green-state surface. For parts over 25 mm in cross-section, data from stereolithography service bureaus indicates that a 0.5° to 1° draft angle on vertical walls reduces peel-induced curl and improves downstream dimensional tolerance.

Recoating behavior is governed by the resin viscosity. Manufacturer-published data place the Brookfield viscosity at 25°C in the 200–400 mPa·s range, which allows high-speed layer deposition on standard wiper blades without heating beyond the rated vat temperature. Processing at 50 μm layer thickness requires lower draw distance and slower blade retraction than the 100 μm build style to prevent air-entrapped regions at the surface of previously cured layers. The green-state flexural modulus is sufficient to support fine pillars and thin ribs above 1.0 mm, but thin walls below 0.8 mm may deflect during peeling unless reinforced with gussets or supported by adjacent geometry. On systems with solid-state 355 nm lasers, the build process uses a photopolymerization threshold determined by the critical exposure of the resin; underexposure produces poor interlayer adhesion, while overexposure increases edge growth and loss of negative feature accuracy.

If Thermal Postcure Is Omitted, the 121°C HDT Claim Does Not Apply

Green parts removed directly from the vat retain significant unreacted monomer and oligomer fractions. The measured heat deflection temperature under ISO 75-2:2013 method B remains in the 50–60°C range for green-state specimens, and water uptake under ISO 62:2008 can exceed the postcure value by a factor of two or more. The target 121°C HDT is obtained only after a forced-air thermal postcure cycle in which the part is ramped from ambient to 121°C at 0.5–1.0°C/min, held for 2–3 h, and cooled slowly to avoid thermal shock. Ovens with uneven airflow can create hot spots above 130°C; at these temperatures the part surface may oxidize and darken before the core reaches full conversion. Therefore the thermal postcure is not a cosmetic secondary operation but the critical step that completes network formation and establishes the water-resistant performance.

Representative datasheet values from manufacturer literature are compiled in the following matrix. Values are not design allowables and must be verified on actual build orientation and postcure oven uniformity.

How Does Water Resistance Differ from Surface Hydrophobicity in Glass-Filled Polypropylene?

Water resistance in ProtoTherm 12120 is evaluated gravimetrically after thermal postcure, not inferred from contact angle alone. A water-repellent thermoplastic surface can remain dimensionally stable under immersion because it does not absorb water, whereas a hydrophilic photopolymer can exhibit low contact angle but low water uptake due to low free volume. Manufacturer-published data under ISO 62:2008 place 24 h water absorption of thermally postcured specimens below 0.60%. This is lower than many unfilled epoxy-acrylate SL resins, which can exceed 1.0% over the same period and show measurable hygroscopic expansion. The practical consequence is that fluid-contact prototypes made from ProtoTherm 12120 maintain clearer dimensional stability in humidity cycling between 20% RH and 80% RH than general-purpose SL resins with otherwise similar tensile modulus.

Quick-connect fluid fittings, water pump impeller prototypes, and sensor enclosures are typical build applications because the material can be machined after postcure and can withstand intermittent water contact at moderate temperature. For parts that will be subjected to continuous pressurized hot water above 60°C, published data for this specific configuration is limited and service validation is required. The material is not a direct replacement for hydrolytically stabilized engineering thermoplastics such as polyphenylene sulfide or polyamide 12 in long-term pressurized service. Supports should be removed before postcure to prevent support remnants from bonding during thermal exposure, and holes smaller than 0.5 mm should be drilled after postcure to avoid resin plugging. Surface roughness from the SLA process can retain water in as-built laminations; sealing or micro-polishing is employed for fluid-contact surfaces where bacterial film accumulation is a concern.

Uncured resin from a drained vat can be filtered through a 100 μm mesh and reused if the material has not exceeded its pot life and has not been contaminated with cleaning alcohol. In production lines, tank-side viscosity drift is tracked with a Brookfield viscometer at 25°C, and the resin is replaced or blended when viscosity exceeds the supplier-specified upper limit. Contamination from loose powder, water, or non-approved cleaning solvents creates observable changes in recoating and scattered sidewall surface roughness. The material should be stored in sealed, light-proof containers between 5°C and 30°C; cold resin exhibits higher viscosity and should be allowed to equilibrate for 12 h before vat transfer.

Differences from Unfilled Epoxy-Acrylate SL Resins and Ceramic-Filled High-Temperature Systems

Three operational boundaries separate ProtoTherm 12120 from adjacent SLA material classes. General-purpose unfilled epoxy-acrylate resins produce smoother sidewalls and faster postcure but exhibit lower HDT and higher water uptake after postcure. Ceramic-filled high-temperature SL resins can exceed 200°C HDT but are more viscous, abrasive to recoater blades, and require thicker support structures due to higher shrinkage stress. ProtoTherm 12120 sits between these classes with a postcure HDT near 121°C, water absorption below 0.60%, and viscosity low enough for standard recoating without vat heating beyond 32°C. In comparison to water-resistant PolyJet photopolymers, the material offers a thermosetting SLA crosslink architecture that does not rely on soluble support removal from deep internal channels; however, PolyJet systems have finer layer thicknesses. The choice therefore depends on whether the dominant risk is dimensional drift from water exposure or layer-step surface finish.

When Fabrication Tolerances Are Tighter Than ±0.1 mm, Thermal Postcure Compensation Must Be Applied Before Machining

Thermal postcure causes a small additional linear shrinkage that is not fully predictable from green-state dimensions. Stereolithography bureaus report that parts scaled uniformly for green-state shrinkage still require local stock allowance of 0.2–0.5 mm on sealing faces and bearing bores before postcure, because thermal postcure shrinkage is anisotropic with respect to build axis and layer orientation. The coefficient of thermal expansion for the cured resin is cited in manufacturer literature as approximately 65–80×10^–6 K^–1 below the glass transition; during the postcure ramp, expansion followed by crosslinking shrinkage can produce non-recoverable deviations if the part is constrained. For this reason, parts are postcured unclamped and ideally supported on a flat ceramic or aluminum plate to allow uniform air access. Holes, slots, and dovetail features that must hold tolerances of ±0.05 mm are machined after postcure rather than built to final size.

Green parts are initially cleaned in a two-stage isopropyl alcohol bath; residual solvent must be fully evaporated before thermal postcure because entrapped alcohol can cause microvoid formation at temperatures above 100°C. A post-clean air-dry period of 2–4 h at ambient or 30°C forced air is employed before ramp-up. Contact with strong alkaline solutions or chlorinated solvents is avoided because these agents can attack the cured network or leave surface residues that interfere with subsequent adhesive bonding. If bonding is specified, abrasion with 240-grit alumina paper followed by air-ionization treatment improves lap-shear repeatability, although published bond-strength data specific to this resin remain limited.

Why Low Water Uptake Is Not Equivalent to Hydrolytic Stability in Continuous Immersion Service

Water absorption and hydrolytic stability are distinct performance attributes. A low 24 h water uptake below 0.60% under ISO 62:2008 indicates limited initial gravimetric moisture uptake, but it does not by itself establish long-term retention of tensile strength and modulus in hot water. The ester and ether linkages in the photopolymer network can undergo hydrolysis at slow rates depending on water temperature and pH. Manufacturer literature positions ProtoTherm 12120 for intermittent water contact and humid environments rather than continuous immersion in boiling water or aggressive aqueous acid. For validation, programs commonly impose 1000 h exposure at 60°C in deionized water and then re-test tensile properties using ISO 527-2:2012. Published data for this specific configuration is limited, so service validation is considered mandatory before replacing a molded engineering thermoplastic.

Postcure Oven Uniformity and Batch-to-Batch Variation

Batch-to-batch variance in thermo-mechanical performance is linked less to resin lot chemistry than to postcure oven loading. On a forced-air oven with a 0.25 m³ working volume, dense loading can create a surface-to-core temperature lag exceeding 10°C during the ramp phase. Operators monitor internal oven air temperature and part surface temperature with a thermocouple attached to a sacrificial witness block, not to the actual part, to prevent local overheating. The witness block is placed at the center of the load and removed after the cycle to measure color change and degree of cure by hardness or FTIR where available. In an actual production environment, parts near the air inlet may darken if the inlet temperature exceeds 135°C; therefore, inlet deflectors are used to distribute air before it contacts the parts.