the part warps

when you fire it

Problem statement

Metal binder jetting — a 3D printing process that jet-prints liquid binder onto metal powder layers, producing "green parts" that must be sintered (heated to near-melting temperature) to fuse into solid metal — promises to be the first additive manufacturing technology economically competitive with casting and machining for medium-volume production runs (hundreds to thousands of parts). Desktop Metal — which went public in a 2020 SPAC merger at a $2.5 billion valuation — bet on this promise, was acquired by Nano Dimension in April 2025, and filed for Chapter 11 bankruptcy in July 2025. Its collapse was financial (the filing cited significant liabilities and liquidity needs), but the technical bottleneck it never fully overcame remains: the sintering step introduces anisotropic, geometry-dependent, and stochastic dimensional distortion that prevents the dimensional precision required for production parts. Parts shrink substantially during sintering as the green body densifies, and the shrinkage varies by direction, wall thickness, local powder density, and proximity to support structures — making final dimensions unpredictable at the tight tolerances industrial customers require.

Why this matters

Metal additive manufacturing is a multibillion-dollar market, but laser powder bed fusion (LPBF) — the dominant technology — has slow build rates and high per-part costs, limiting it largely to aerospace and medical applications where geometry complexity justifies the cost. Binder jetting prints substantially faster than LPBF and can use lower-cost MIM-grade metal powders rather than the gas-atomized powders LPBF requires, making it potentially cost-competitive with casting for production quantities. If the sintering distortion problem were solved, metal additive manufacturing could expand from a niche market into mainstream manufacturing for automotive, consumer electronics, and industrial components.

What’s been tried and why it hasn’t worked

Desktop Metal and ExOne (acquired by Desktop Metal in 2021) developed proprietary simulation software to predict sintering shrinkage and pre-compensate part geometry. However, the simulations rely on material models calibrated to specific powder lots, and real-world powder characteristics vary batch-to-batch (particle size distribution, morphology, oxide content). The sintering process itself is highly nonlinear: shrinkage begins at grain boundaries, proceeds through bulk diffusion, and is influenced by local packing density, binder burnout residues, and thermal gradients within the furnace. Support structure design affects local constraint and distortion but the interactions are complex and geometry-specific. Post-sintering machining can correct distortion but adds cost and defeats the near-net-shape advantage of additive manufacturing. HP's Metal Jet — a competing binder jetting technology unveiled in 2018 — did not reach general commercial availability until the Metal Jet S100 launch in September 2022, and as a binder-jet system it depends on the same debind-and-sinter densification step. Markforged (whose Metal X uses bound-metal extrusion plus the same debind-and-sinter step) was acquired by Nano Dimension in April 2025 for $116 million.

What would unlock progress

Physics-informed machine learning models trained on large datasets of sintered part geometries (scan of green part vs. scan of sintered part) could learn the complex, nonlinear mapping between as-printed geometry and final dimensions, enabling accurate pre-compensation without requiring first-principles sintering simulation. In-situ monitoring during sintering (using dilatometry, optical scanning, or acoustic emission) could enable real-time process adjustment. Novel sintering approaches — pressure-assisted sintering (SPS/FAST), microwave sintering, or two-stage sintering profiles — could reduce the temperature and time at which densification occurs, narrowing the distortion window. Standardized characterization protocols for powder feedstock that predict sintering behavior (beyond current particle size distribution specs) would reduce lot-to-lot variability.

Entry points for student teams

The cheapest real door into the sintering step is metal-filled filament. BASF Forward AM's Ultrafuse 316L prints green parts on an ordinary open-material desktop FFF printer, and Forward AM's named partners (DSH Technologies, Elnik Systems) run the debind-and-sinter cycle as a paid service — so a team can design a benchmark set varying wall thickness, overhang angle, aspect ratio, and support contact, measure every feature green, send the parts out, and characterize the anisotropic shrinkage that comes back. This is not binder jetting, and a writeup should say so, but it is the same MIM-derived debind-and-sinter densification step where the distortion originates. A team with no printer works the modeling side instead: build a continuum sintering model in MOOSE (open-source multiphysics, LGPL-2.1) and calibrate it against the published green-versus-sintered dimensional measurements in Sadeghi Borujeni et al. (2022), then test how much of the reported distortion a geometry-feature regression predicts without any first-principles sintering physics at all. A production binder jetting machine is a third route but an access line rather than an entry point: an ExOne Innovent+ with a matched debinding and sintering furnace is a six-figure capital installation owned by specific powder-metallurgy and AM research groups, so a team takes that route only when a particular faculty lab has already committed the machine time. Relevant disciplines: materials science, mechanical engineering, manufacturing engineering, data science.