Portfolio
Sustainable AM / Energy-Efficient Ceramic AM (REAP)
Sustainable Manufacturing · Ceramics

A ceramic 3D printer that burns its own kiln.

REAP couples photothermal direct ink writing with a self-propagating exothermic reaction. A 10-second ignition releases a 2,000 °C front that consolidates a complex preform at ~100× lower energy and three orders of magnitude lower CO2 than conventional sintering.

  • Reactive AM
  • Self-propagating high-T synthesis
  • Polymer-derived ceramics
  • COMSOL
  • LCA
Energy
~100× less
CO2
2.23 kg/kg ceramic
Consolidation
~5 s
Yield strength
180 MPa

J. Cleaner Production 452, 142122 · 2024 Published · Patent Pending

REAP process flow: reactive elemental powder mixture combined with preceramic oligomers extruded under UV light to print a green preform, then ignited so a reaction front sweeps through the part, with five thermal imaging frames showing the 2000 degree Celsius front propagating across a honeycomb lattice

Why it matters

Global ceramic production consumes >182 TWh/year, with high-temperature firing accounting for ~75% of energy cost. Industrial heating drives roughly half of US greenhouse-gas emissions. Conventional kilns hold parts at >1,000 °C for hours to days; binder jetting and SLA still need lengthy post-sintering.

Meanwhile, demand keeps rising for geometrically complex ceramics — gyroid heat exchangers, refractory lattices, hypersonics components — that pressureless sintering cannot produce. The sector needs a process that breaks the energy–geometry trade-off and decarbonizes a hard-to-abate industry at the same time.

How it works

  • Step 1 — Photothermal printing. A reactive ink (Ti:Si:C 3:1:2 with SMP10 polycarbosilane) extrudes through a 1 mm nozzle with co-located UV at 9.6–11.4 W/cm² for in-situ cure — no photoinitiator, tolerates heavy opaque loading.
  • Step 2 — Self-sustaining ceramization. A 192 W tungsten basket ignites one spot for ~10 s. Ti + C → TiC and 3Ti + Si + 2C → Ti3SiC2 release a 2,000 °C front that self-propagates at ~130 cm/min with no further external energy.
  • Energy decoupling. Because the heat source is intrinsic, external draw is decoupled from part size — per-cm³ energy falls as parts grow (192 J/cm³ at 10 cm³ → 19.2 J/cm³ at 100 cm³).
  • Characterization. Time-resolved XRD + Rietveld for phase evolution; SEM microstructure; ASTM C1161 strength + C1421 fracture toughness; DSC + FTIR; COMSOL thermal sim (MAPE 2.6% vs. experiment).
  • Lifecycle analysis. End-to-end CO2-eq footprint benchmarked against binder jetting and SLA equivalents.
Six-panel composite: reactive ink viscosity versus shear rate compared with as-received SMP10 and heat-treated oligomer, four photographs of printed green preforms including a honeycomb, hex blocks, Olympic rings, and a scaffold, the live UV photothermal nozzle, temperature vs time curves for 64 to 74 weight percent preceramic content, FTIR spectra showing Si-C=C and Si-H2 evolution after cure, and a UV power versus deposition speed process map with degree-of-cure contours
Print envelope Reactive ink rheology, cured green preforms across geometries, FTIR cure verification, and the UV-power × deposition-speed window where degree-of-cure stays above 0.6.
Reaction front validation: IR thermal images of the printed sample at 0, 0.5, 1.5, and 2.5 seconds showing a bright propagating front, the matching COMSOL simulation snapshots in the same time intervals, an experimental versus simulation temperature curve that peaks near 2200 degrees Celsius, and a plot of reaction front temperature and propagation speed across preceramic content from 5 to 25 weight percent
Reaction front Live IR images vs. COMSOL agree on a ~2,000 °C front sweeping the preform in <3 s. Propagation speed stays near 130 cm/min while front temperature falls with higher preceramic loading.

What we found

192 J/cm³ Energy / 10 cm³ part ~100× less than conventional sintering; falls to 19.2 J/cm³ at 100 cm³.
2.23 kg/kg CO2-eq footprint vs. 3,086 (binder jet) and 616 (SLA) — three orders of magnitude lower.
1.7 MPa·m1/2 Fracture toughness Yield strength 180 MPa; porosity 25.6% (60% lower than prior reactive AM).
6.5 °C ΔT Gyroid HX demonstrator 1.5 cm unit cell; leak-free across the tortuous internal topology.
Two-panel benchmark: left panel is a log-log scatter of energy consumption versus consolidation time clustering conventional sintering, flash sintering, laser sintering, SPS, microwave sintering, and fast firing, with REAP marked as a red ellipse far below all of them at 10 cm cubed and 100 cm cubed specimen sizes; right panel is a stacked bar chart of cradle-to-gate CO2 footprint with binder jetting at 3086 kg CO2 per kg, SLA at 616, and REAP at 2.23 with the small REAP bar broken out at log scale
Benchmark Energy–time map (left): REAP sits one to two orders of magnitude below every prior sintering route. CO2 footprint (right): three orders below binder jetting, two below SLA.
Phase evolution figure: stacked XRD patterns at four time points t1 through t4 showing Ti and Si peaks giving way to TiC and ultimately Ti3SiC2 reflections, the corresponding temperature-time trajectory peaking near 2200 degrees Celsius, and two SEM micrographs at 100 micron scale contrasting a rough porous green microstructure with a denser ceramic microstructure
Phase evolution Time-resolved XRD traces Ti and Si peaks collapsing into TiC and Ti3SiC2 as the front passes. SEM shows the porous green preform consolidating into a dense MAX-phase ceramic.
Gyroid ceramic heat exchanger demonstrator: photograph of a black gyroid block alongside a millimeter ruler showing 1.5 cm unit cell, dark-field optical micrograph of the wall at 250 micrometer scale, hot-in hot-out cold-in cold-out temperature traces over 20 seconds, and a labeled IR thermal image of the operating exchanger with hot in arrow on left and cold out arrow on top
Demonstrator 1.5 cm gyroid unit cell printed and reacted in REAP. Steady-state operation gives a 6.5 °C hot-side drop across leak-free internal channels.

So what. Decarbonizes a hard-to-abate sector while unlocking gyroid lattice heat exchangers, hypersonic refractories, and MAX-phase components that traditional ceramic processing cannot produce — with seconds-scale throughput and an efficiency advantage that grows with part size.

Paper & resources

Journal article J. Cleaner Production 452, 142122 (2024)
DOI 10.1016/j.jclepro.2024.142122
Patent Pending
Author Prashant Dhakal
BibTeX · JCP 2024 
@article{liu2024reap,
  title   = {Energy-efficient rapid additive manufacturing of complex geometry ceramics},
  author  = {Liu, Ruochen and Hou, Aolin and Dhakal, Prashant and Gao, Chongjie and Qiu, Jingjing and Wang, Shiren},
  journal = {Journal of Cleaner Production}, volume = {452}, pages = {142122},
  year    = {2024}, doi = {10.1016/j.jclepro.2024.142122}
}