Epoxy embedded 3D printing has produced a recognisable Benchy, thanks to a home-built support bath printer constructed by [Riley] of Hackaday-featured maker Riley’s Lab. The rig submerges the entire print in a shear-thinning support gel, holding extruded liquids in place until they cure, a technique that sidesteps one of the most stubborn constraints in unconventional extrusion.
Conventional fused-deposition printing works because the material behaves itself: it flows through a hot nozzle and then stays where it lands. Plenty of other interesting substances simply won’t do that. Too low a viscosity and the extruded bead slumps and merges before it can set, making layered geometry impossible. The embedded printing approach solves this by surrounding the workpiece with a bath that is simultaneously supportive and penetrable.
Building the Support Bath and Syringe Pump
Rather than a standard extruder, [Riley] fitted his machine with a largely 3D-printed syringe pump in the tool carriage. A syringe pump applies steady, controllable pressure to almost any liquid, which immediately opens the door to materials that would destroy or clog a conventional hotend. The support bath itself is a mixture of mineral and vegetable oil, with fumed silica added to produce the required shear-thinning behaviour. In plain terms, the gel holds firm when undisturbed (keeping freshly deposited material in position) but yields and flows the moment the extrusion needle moves through it, sealing back up behind the needle’s path.
Before committing to anything exotic, [Riley] validated the setup with cream cheese. It is cheap, cooperative under a syringe, and retains its shape well after extrusion, genuinely useful properties for a first functional check. The resulting tardigrade model confirmed that the motion system and pump were working correctly.
Epoxy Embedded 3D Printing Outlasts the Silicone Attempt
The first serious material test used Dow Sylgard 184, a two-part polydimethylsiloxane silicone. In the bath it looked promising, bar some stringing between features, but the inter-layer adhesion turned out to be poor. When [Riley] attempted to remove the cured print, it disintegrated. The bath had done its job; the silicone simply hadn’t bonded to itself sufficiently between passes.
The second attempt switched to a two-part epoxy, and the results were considerably better. Stringing was still present, but the print held together as the bath was washed away in isopropyl alcohol. The finished epoxy Benchy showed decent shock resistance, which, for an unconventional print in an experimental rig, is a meaningful outcome. The support bath had contained the liquid epoxy reliably through the curing period, and the wash-out process left a coherent, handleable object.
The shear-thinning behaviour of the oil-and-fumed-silica bath is the same fundamental principle driving research in more specialised directions. According to PatSnap Eureka, Regenovo specialises in bioink development for 3D bioprinting using shear-thinning hydrocolloid systems, employing multi-component blends that combine sodium alginate, gelatin methacryloyl (GelMA), and hyaluronic acid derivatives. Where Riley’s bath uses oils and fumed silica to achieve the same yield-stress behaviour, bioprinting researchers are engineering hydrocolloid chemistries with biocompatibility requirements layered on top. The underlying physics (a material that flows under shear and holds firm at rest) is common to both approaches.
Epoxy embedded 3D printing in a home workshop, using a printed syringe pump and a bath mixed from off-the-shelf oils, is a tangible demonstration of how accessible this technique has become. Commercial and research implementations of embedded printing have been covered previously; seeing an independent builder iterate through material failures and reach a shock-resistant result with standard two-part epoxy shows the method is workable well outside a laboratory context. [Riley] has already identified stringing as the next problem to address, which gives this project a clear direction forward.
