How Does Burnout Kinetics Govern Cationically Cured Epoxy Pattern Removal in Shell Investment Casting?

The burnout behaviour of Somos 8120 in shell investment casting is governed by the interplay of shell permeability, heating ramp rate, and oxidative decomposition timing; patterns are produced in a 355 nm galvanometer-scanned stereolithography system at 0.05 mm or 0.1 mm layer thickness, with interior drainage passages of at least 2.0 mm and a cleaning sequence using isopropanol or tripropylene glycol monomethyl ether to remove unsolidified resin. The manufacturer’s reported liquid-state viscosity at 30 °C, below 300 mPa·s, permits gravity-assisted drainage from wall sections as thin as 1.2 mm. Shell construction uses a colloidal silica primary slurry with 30–35 % silica solids and 70–80 % zircon stucco of 0.1–0.2 mm mesh, followed by fused silica intermediate coats; each coat is dried at 22–25 °C and 40–50 % relative humidity until stable mass is recorded. The burnout programme starts with a 1–2 °C/min ramp to 150–180 °C and a 1–2 h hold to maintain shell internal pressure below 20–50 kPa, followed by a 3–5 °C/min ramp to 650–750 °C in an oxygen-containing atmosphere; the upper temperature is selected according to the alloy group and shell face coat. Residual ash is measured with ASTM D482 or ISO 3451-1; any deviation beyond the foundry’s stated ash ceiling triggers borescope inspection of the cavity and compressed-air evacuation before metal pour. Patterns are pre-inspected using general dimensional tolerances from ISO 2768-1; z-axis anisotropic shrinkage during 0.1 mm layer builds is compensated by a scaling factor of 1.005–1.010. Terminal castings include aluminium A356 pump housings, stainless steel valve bodies, and cobalt-chromium impellers; after shell preheat at 500–600 °C, metal is poured at controlled velocity to minimise turbulence and shell inclusions. This process is restricted to alloys whose pour temperature does not destabilise the shell during the high-temperature segment; published data for the specific oxidation rate of Somos 8120 in chemically bonded shells is limited, so each foundry validates the ash residue and shell cracking threshold on a representative pattern before production lots.

Silicone tool fabrication from a Somos 8120 master pattern begins with a barrier coat or release-agent application before platinum-catalysed addition-cure silicone is cast, because residual photoacid from cationic epoxy polymerisation can inhibit the hydrosilylation crosslinking reaction at the pattern interface. A two-part platinum silicone is mixed at a 10:1 base:catalyst ratio by weight, degassed under vacuum at 2–5 mbar until gas evolution ceases, and poured within 10–15 min of mixing; the filled mould is then vacuum-degassed again at 1–5 mbar for 2–5 min to remove air from rib features and surface lettering. Curing is carried out at 60–70 °C for 2–4 h, remaining below the heat deflection threshold of the epoxy master to avoid progressive distortion during the first mould cycle. When the final cast polyurethane requires a non-optical surface finish, the master is polished with 600-grit wet abrasive to a surface roughness of Ra 0.4 µm prior to mould casting; straight-pull sections retain draft angles of 0.5°–1.0°, while undercut regions are handled with a two-part block mould and methyl methacrylate registration frames. The cured silicone cavity is then used for vacuum casting of polyurethane prototypes; typical prepolymer:hardener ratios are 1:1 or 2:1 by weight, and mineral or short glass-fibre fillers are added at 20–50 % of resin weight in rigid or fire-retardant grades. Components produced in this downstream step include automotive intake manifold prototypes, handheld medical device housings, and elastomeric gasket mock-ups; a single silicone cavity commonly produces 20–40 polyurethane components before parting-line tearing or surface degradation reduces dimensional fidelity. Dimensional checks of the cast polyurethane are referenced to ISO 2768-1 general tolerances, and the silicone mould itself is inspected for hardness shift using ISO 48-4; if Shore A hardness falls by more than 5 points from the initial 40–70 Shore A range, the mould is retired from dimensional-critical work.

When pressure tap arrays and balance inserts are integrated into moderately loaded wind tunnel test sections

Wind tunnel sections fabricated from Somos 8120 are not consumed as burnout patterns; they are finished, instrumented, and repeatedly loaded. Stereolithography build parameters at 0.1 mm layer thickness are selected to maintain a surface step height below 0.01 mm after sanding, and internal pressure galleries are designed with teardrop cross sections and a minimum bore of 1.0 mm to permit solvent drainage. After the build, the model is washed in two successive solvent baths, then post-cured in a UV chamber at 60 °C for 60–90 min until co-printed tensile coupons tested under ASTM D638-14 reach stable tensile strength. Pressure tap orifices of 0.3–0.6 mm are drilled perpendicular to the flow surface; stainless steel or polyether ether ketone tubes of 0.5–1.0 mm inner diameter are bonded with a two-part room-temperature epoxy adhesive mixed at 2:1 by volume, with fillets cured at 22–25 °C for 24 h. Surface discontinuities around the taps are levelled with polyester finishing filler and sanded to a step below 0.01 mm to avoid premature boundary layer transition. Balance and support loads are transferred through aluminium or stainless steel spigots bonded or cast into the root section; static flexural checks are performed according to ISO 178 to ensure model deflection remains below facility blockage tolerance. The maximum aerodynamic load for this unfilled photopolymer is limited to low-speed and moderate dynamic pressure conditions; published data for specific high-speed configurations is limited, and any campaign above approximately Mach 0.3 requires a facility-specific load–deflection validation. Terminal articles include wing-body research models, turbine blade cascade test pieces, and air intake swirl probes; each model is surface-sealed with an epoxy coating and dimensionally checked with structured-light scanning to ±0.15 mm per 200 mm.

For short-run thermoforming and vacuum-forming tools produced directly from stereolithography, the principal process conflict is not print resolution but the heat deflection limit of the unfilled epoxy network under contact with hot sheet. The Somos 8120 tool shell is built at 0.05 mm layer thickness for shallow draw surfaces, then backfilled with a room-temperature aluminium-filled epoxy mixed at 4:1 by weight resin:hardener to provide stiffness and thermal mass; the backfill is poured in layers no thicker than 10 mm to avoid exothermic temperatures above 60 °C near the shell. The tool surface is sealed with a two-part epoxy coating mixed at 2:1 by volume and lightly sanded to remove gloss. During vacuum-forming, the tool surface is preheated to only 35–50 °C, and the polyethylene terephthalate glycol or ABS sheet is heated to 160–200 °C but is not held against the tool for more than 3–5 s per cycle; the tool is then cooled with compressed air through printed internal channels. Published data for this specific configuration is limited; before production, the tool is cycled for 20 h without sheet to confirm that surface temperature at the photopolymer backfill interface does not exceed 50 °C. The downstream parts are packaging trays, shallow automotive interior trim skins, and lens cover prototypes; dimensional conformance is verified with ISO 2768-1, and surface quality is assessed under ISO 4287 roughness parameters with an Ra target no greater than 0.8 µm after 50 cycles.

Service life boundaries for low-thermal-load electronic enclosure prototypes with snap-fit geometry

Electronic enclosure prototypes built from Somos 8120 are used in dry, low-thermal-load validation programmes where flammability rating is either not required or supplied by a post-applied coating. Snap-fit arms and cantilever latches are printed at 0.05 mm layer thickness with orientation arranged so that the layer planes are not parallel to the primary tensile bending stress; the anisotropic tensile strength of stereolithography parts can reduce cantilever root capacity by 15–30 % compared with the vendor’s x-y plane datasheet value. Tensile modulus and elongation at break are measured on co-printed coupons according to ASTM D638-14; the allowable snap-fit strain is derated by a factor of 2.5 from the measured ultimate elongation to compensate for moisture absorption. Water absorption is evaluated under ASTM D570-22 at 23 °C and 50 % relative humidity; if the enclosure is intended for service above 60 % relative humidity, a two-part epoxy or UV-cured conformal coating of 50–100 µm dry film thickness is applied at a 4:1 volume mix ratio to reduce dimensional swelling. Threaded inserts are installed with cyanoacrylate adhesive and pull-out tested at a crosshead speed of 1.0 mm/min; if failure occurs below the design pull-out load, the boss is thickened or an elastically averaged mount is used. The photopolymer is not assumed to meet any UL 94 classification; when a V-2 or better rating is required, a flame-retardant intumescent coating is applied and tested on the actual printed substrate under UL 94 vertical burn criteria. Terminal uses include display bezel mock-ups, connector shell fit-check fixtures, and handheld diagnostic device housings used at 10–35 °C in dry indoor environments; snap-fit datum features are verified before assembly by contact gauging to ±0.05 mm.

Anatomical models fabricated from Somos 8120 for orthognathic and maxillofacial surgical planning are generated from computed tomography DICOM data with slice spacing of 0.3–0.6 mm and a segmentation threshold selected by the radiology workstation for cortical bone. The stereolithography build is performed at 0.1 mm layer thickness, with model orientation adjusted to minimise stair-step artefact on the inferior alveolar nerve canal and tooth roots; internal segmented voids are drained through 2.0 mm minimum openings and washed in the same solvent sequence used for investment casting patterns to prevent residual liquid from leaching later. Post-curing is conducted at 60 °C for 60–90 min until the surface is tack-free to gloved handling. The model is sealed with a two-part epoxy surface coating mixed at 2:1 by volume A:B to reduce water uptake during wet-lab simulation; if model contact with sterilised instruments is anticipated, the coating is disinfected with 70 % ethanol/30 % water by volume, but no implant or prolonged tissue-contact classification is implied. Dimensional accuracy of the printed model is verified against the DICOM reference with structured-light scanning using ISO 2768-1 general tolerances; deviations at anatomical landmarks are retained in a validation record. Terminal uses include osteotomy cutting guides used only as visual and spatial references in simulation, patient-specific drilling templates produced after vacuum-formed transfer shells, and teaching casts for implant trajectory planning. Published data for the long-term sterility and mechanical ageing of this photopolymer under repeated disinfectant exposure is limited; medical device manufacturers assess each printed model under their own quality system before clinical use.