Binder Jetting for Metal Parts: Process Window, Shrinkage, and Economics

Binder jetting is the only metal AM process designed from the ground up for volume production economics. It has no laser or electron beam — the printing step is cold, fast, and cheap. The physics of densification happens later, in a sintering furnace. That separation of printing from densification is both the process's greatest advantage and its primary engineering challenge.

1. Process steps

Metal binder jetting proceeds in five steps. Understanding what each step does is essential for anticipating where problems occur.

Step 1 — Print (green body formation): A thin layer of metal powder (typical D50: 5–30 µm) is spread across a build platform. An inkjet printhead deposits a liquid binder in the cross-section of each layer. The binder saturation (volume of binder relative to powder void space) is typically 60–100%. Layer thickness: 50–200 µm. Build speed: 10–100× faster than LPBF because there is no melting.

Step 2 — Cure: The completed green body is heated in the build box to 150–200°C to evaporate solvent and crosslink the binder polymer. This strengthens the green body enough to handle.

Step 3 — Depowder: Loose unfused powder is removed from around the green body. Compressed air jets, brushes, and vibration are used. Critically, no support structures are needed — loose powder supports the part during printing, and the powder bed supports the green body during depowdering. This is a key economic differentiator from LPBF.

Step 4 — Debind: The binder must be removed before sintering, without destroying the green body. Two main approaches:

• Catalytic debind (ExOne, Digital Metal): Nitric acid vapour catalytically depolymerises the binder at 100–120°C, removing it progressively from the outside in without a liquid phase. Fast (4–12 hours).
• Thermal debind: The binder burns off during initial furnace heating (< 600°C ramp) as part of the sintering cycle. Simpler but requires careful atmosphere management to prevent oxidation.

The post-debind part is called the "brown body" — it holds its shape from powder-particle friction and residual binder, but is very fragile.

Step 5 — Sinter: The brown body is sintered in a furnace at 60–90% of the alloy's melting temperature. Solid-state diffusion bonds particles and eliminates porosity. For 316L, sintering at 1360–1380°C for 2–4 hours achieves 97–99.5% relative density. During sintering, the part shrinks uniformly as pore volume decreases.

The entire cycle — print to sintered part — takes 24–72 hours depending on part size and debind method.

2. The shrinkage challenge

Sintering shrinkage is the defining accuracy challenge of metal binder jetting.

Magnitude: Linear shrinkage is typically 17–22% in all three dimensions for common alloys. A 100 mm green body becomes approximately 79–83 mm after sintering. For 316L with optimised D50 powder (~22 µm) and a well-controlled sintering cycle: ~20% linear shrinkage is the design baseline.

Why it matters: The final sintered dimension must meet drawing tolerances. Pre-scaling the print file to compensate is standard practice — if the target dimension is 100 mm with a shrinkage factor of 1.200, the print file is scaled to 120 mm.

Isotropy: Shrinkage is isotropic (equal in X, Y, Z) for fine, spherical powders with uniform packing density. This is achievable in practice. The critical caveat:

• Coarser powder (D50 > 30 µm) and/or poor flowability → non-uniform packing → anisotropic shrinkage → different scale factors required for X/Y vs. Z
• Part geometry affects local shrinkage: thin sections, features with low packing density, and overhanging regions sinter at slightly different rates than bulk sections
• Complex internal features (channels, cavities) experience sintering differently than solid regions — the cavity constrains shrinkage of the surrounding wall

Practical dimensional accuracy: With optimised process, sintering fixture use, and calibrated scale factors:

• Typical tolerance: ±0.3–0.5% of nominal dimension
• Best-case with post-sinter machining: ±0.05 mm on critical surfaces

For comparison, LPBF achieves ±0.1–0.2% as-built. Binder jetting as-sintered is less precise, but the economics often justify adding a light machining pass on critical surfaces.

Sintering fixtures: Complex geometries with thin walls or tall features require ceramic setters or support structures placed in the furnace to prevent gravity-induced slump during sintering (the brown body becomes plastic before densification is complete). Fixture design is a process engineering task, not automated.

3. Supported materials

Most mature:

Challenges:

• Aluminium: Significant oxide layer on Al particles inhibits sintering densification. High vapour pressure of Al at sintering temperatures and strong oxide stability make binder jetting of aluminium difficult. No commercially mature offering as of 2025.
• Titanium: Sintering titanium requires very low oxygen partial pressure (< 10 ppm) to prevent embrittlement. Achievable with hydrogen-atmosphere or vacuum sintering furnaces, but adds complexity and cost. Limited commercial availability.
• Reactive alloys generally: Any alloy that forms stable oxides readily (Ti, Al, reactive Ni alloys) needs carefully controlled atmosphere. EBM is often more practical for Ti.

4. Density and mechanical properties

Achievable sintered density determines mechanical properties. At 97–99.5% relative density, mechanical properties of binder jetted 316L are comparable to wrought annealed material in most respects:

BJ 316L has lower yield strength and hardness than LPBF because sintering produces equiaxed, fully annealed grains without the cold work and rapid solidification effects of LPBF. It has comparable or better ductility and toughness. For most structural applications, BJ 316L meets design requirements.

Microstructure: BJ sintered microstructure is equiaxed with no crystallographic texture — isotropic properties in all directions. LPBF has columnar texture and some directional property variation. For applications where isotropy matters (pressure vessels, complex load paths), BJ has an advantage.

Residual stress: Near zero. The sintering temperature is well above stress relief temperature for most alloys.

5. Surface finish

As-sintered surface finish is better than LPBF as-built on most surfaces:

The better BJ surface finish occurs because the powder is not melted during printing — no spattering, no laser-induced surface ripple, no condensate redeposition. The sintering process smooths the surface slightly as necks grow between particles.

However, binder jetting has a floor set by the powder D50: surface features smaller than ~3× the powder D50 are not reliably resolved. For D50 = 22 µm powder, minimum resolved feature ≈ 60–70 µm.

6. Economics vs. LPBF

This is where binder jetting makes its case.

The key structural advantages over LPBF:

1. No support structures — loose powder supports the part during printing. Parts are nested in 3D within the build box at very high packing efficiency (40–60% volumetric utilisation). LPBF packing utilisation for supported parts is typically 10–30%.

2. Fast print speed — inkjet deposition is 10–100× faster than laser scanning per unit area. A 1-litre build box can be filled in hours rather than days.

3. Cheap print consumable — binder is inexpensive. There is no laser source requiring maintenance or replacement.

4. Sintering is batch-scalable — once parts are printed and debound, the sintering furnace is not a bottleneck. Multiple print-box runs can queue through a single furnace.

Break-even analysis vs. LPBF:

The cross-over point is geometry and volume dependent. Approximate guideline for SS316L brackets/structural components:

These numbers assume medium complexity parts (30–100 cm³). Simple geometries favour MIM at high volume; complex geometries with low volume favour LPBF.

LPBF retains the advantage when:

• Part requires < 5 identical pieces
• Design is still changing (LPBF has faster iteration cycle — no sintering tooling or sintering programme to develop)
• Material is outside BJ's qualified range (Ti, Al)
• Dimensional tolerance is ±0.05 mm or tighter on critical surfaces (LPBF as-built, or LPBF + light machining)
• Fatigue criticality requires LPBF + HIP microstructure control

7. Equipment landscape (2024)

HP Metal Jet S100 is the 2024 high-water mark for production binder jetting quality and throughput. It uses a water-based binder with thermal curing, runs at 630 litres/hour, and is commercially qualified for 316L and 17-4PH to automotive and industrial standards.

8. When binder jetting fails

BJ has real limitations that should not be papered over:

Complex internal channels: Sintering a part with internal channels imposes asymmetric shrinkage constraints. The channel wall is constrained differently from the bulk material — the two can sinter at different rates, causing channel distortion or cracking. Channels below 3 mm diameter with L/D > 10 are high-risk in BJ. LPBF + abrasive flow machining is often more reliable.

Thin walls < 1.5 mm: Green body strength is low. Thin walls may crack or distort during depowdering or sintering. Minimum reliable wall: 1.0–1.5 mm, aspect ratio < 20:1 without sintering fixtures.

Fatigue-critical parts: BJ porosity of 0.5–3% (for non-optimised processes) or 0.5–1% (for optimised) is mostly spherical — less damaging than LOF porosity but still present. Grain boundary contamination from binder residue can occur if debind is incomplete. For fatigue-critical aerospace or medical parts, LPBF + HIP remains the more qualified route with more accumulated data.

Tight tolerances on multiple features: Each feature shrinks by the same percentage, but the absolute tolerance in mm scales with feature size. For a 10 mm bore: ±0.5% = ±0.05 mm — acceptable. For a 0.5 mm wall: ±0.5% = ±0.0025 mm — technically acceptable but near the edge of sintering repeatability. Multiple mating features with tight stack-up are difficult.

Decision framework

Further reading

• Process selector — compare binder jetting against LPBF, SLS, and MIM for your part requirements
• Cost per part tool — model LPBF vs. BJ cost including build rate, nesting, and post-processing
• Shrinkage compensation tool — calculate pre-scaled print dimensions for a target sintered size
• Dimensional accuracy tool — assess achievable tolerances by process and feature type
• Packing density tool — estimate how many parts fit in a build box at a given packing fraction