Additive + Subtractive

Binder Jet Metal 3D Printing: 5 Essential Steps for Fabrication Shops

Binder jet metal 3D printing lets a shop build metal parts without a laser and without a melt pool. This guide is for engineers and shop owners who want to know how the process actually runs, where the tolerances come from, and when a CNC machine is still the better call. Read it and you can judge whether a part belongs in binder jet or on a mill.

Get Instant Quote Talk to an Engineer

No laserSinter shrink 15–20%Batch buildCNC finishing

binder jet metal 3D printing build plate with green parts

On this page

1
Key takeaways
2
How binder jet metal 3D printing builds a part
3
Why shrink compensation decides your tolerances
4
Where binder jet metal 3D printing goes wrong
5
Pairing binder jet with CNC finishing at GreatLight
6
5 essential steps from CAD file to finished part
7
Binder jet vs CNC machining: which to pick

Quick answer

Key takeaways

It is a sintering process, not a melting processA print head deposits binder into metal powder; the part is cured, de-powdered, then sintered in a furnace.

Shrinkage is the whole gameGreen parts shrink 15–20% linearly during sintering, so the CAD model must be scaled up before printing.

Powder removal limits the designInternal channels need two open ends and a diameter of at least 3–5 mm, or trapped powder stays inside.

Sintered surfaces are roughExpect Ra 6–10 μm as-sintered. Critical faces need CNC finishing to reach Ra 0.8–1.6 μm.

Best for batches, not for one-offsA full build plate amortizes the furnace cycle. A single part is usually cheaper on a 3-axis mill.

How it works

How binder jet metal 3D printing builds a part

Binder jet works like a powder-bed inkjet printer. A recoater spreads a thin layer of metal powder, typically 30–60 μm thick. A print head then jets a liquid binder into the areas that will become the part. The build plate drops by one layer thickness, fresh powder spreads on top, and the cycle repeats until the build is complete.

The result is not a metal part yet. It is a green part: metal particles glued together by cured polymer. It has roughly the strength of chalk and will break if you handle it carelessly. Binder content is usually 1–3% by weight, and the green part is porous and fragile. Every downstream step exists to turn that fragile shape into dense metal.

This is where binder jet diverges from laser powder-bed fusion. There is no melt pool, so there are no residual stresses from rapid solidification. Parts do not need support structures for thermal reasons, and you can stack them tightly in the build box. That packing density is what makes the process economical at volume.

1
Layer thickness30–60 μm typical; thinner layers give better surface finish but slower builds.
2
Binder content1–3% by weight; burned out during sintering.
3
Green strengthLow. Handle with padded trays and soft tooling only.

Sintering science

Why shrink compensation decides your tolerances

During sintering, the part is heated to 80–90% of the alloy melting point in a controlled atmosphere. The binder burns off first, then the metal particles bond together by solid-state diffusion. The part becomes dense, typically 97–99% of theoretical density, and it shrinks.

Linear shrinkage for common alloys runs about 15–20%. A 100 mm dimension in CAD can come out at 82–85 mm after sintering. The exact number depends on alloy, powder particle size distribution, binder loading, and furnace profile. It is not a constant you can copy from a datasheet.

So the CAD model is scaled up by a compensation factor before printing. If the actual shrink differs from the assumption by even 0.3%, a 100 mm feature moves 0.3 mm. That is 60 times looser than a ±0.005 mm CNC tolerance. For this reason, binder jet holds loose tolerances as-sintered and relies on machining for anything critical.

1
Density target97–99% of theoretical; residual porosity affects fatigue life.
2
Shrink window15–20% linear, alloy dependent. Verify with a test coupon per batch.
3
As-sintered toleranceRoughly ±0.5% of dimension. Tight features need machining.

Limits and pitfalls

Where binder jet metal 3D printing goes wrong

The most common failure is a shrink factor that was assumed instead of measured. Shops print a full build, sinter it, and find every dimension off by 0.4 mm. The fix is cheap: print a coupon with the same alloy and binder batch, sinter it alongside the real parts, and measure before you commit to a production run.

The second failure is trapped powder. A channel with one closed end keeps loose powder inside. During sintering that powder sinters into a hard plug you cannot remove. Design with two open ends, or plan to drill the escape hole after sintering. For blind features, keep depth under 3 times the width.

The third failure is warping. Long thin sections distort because the shrink is not uniform in every direction. Add ribs, keep wall thickness consistent, and orient the longest dimension flat. If a part still warps, it usually needs a fixture during sintering or a design change rather than a furnace tweak.

1
Always run a shrink couponOne coupon per alloy and binder batch. It costs almost nothing.
2
Give powder a way outTwo open ends on every internal channel, 3–5 mm minimum diameter.
3
Keep walls uniformSudden thickness changes cause differential shrink and cracking.

Hybrid approach

Pairing binder jet with CNC finishing at GreatLight

Binder jet is a near-net process, not a net process. Treat the sintered part as a blank, the same way you would treat a casting. The value comes from letting the additive step create geometry that would be expensive to machine, then letting the CNC step create the dimensions that actually seal, fit, and wear.

GreatLight runs 127 high-precision CNC machines, including 16 simultaneous 5-axis machining centers, 12 four-axis mills, 27 three-axis machines, and 16 mill-turn centers. Maximum processing size is 4,000 mm, with travels of 4,000 × 400 × 150 mm on the large frame and 750 × 1,150 × 550 mm on the medium frame. Sintered blanks fit the same workholding as castings.

For a hybrid job, we ask for the CAD model, the alloy, the faces that need tight tolerance, and the batch quantity. Quotation and free DFM analysis come back within 12 hours. Production can start within 24 hours, and parts ship in 3–5 days. We hold ±0.005 mm on machined features and inspect 100% before shipment.

1
Send the model and the critical facesWe mark up which surfaces go to the mill and which stay as-sintered.
2
Batch quantity mattersVolume builds amortize the furnace cycle better than single parts.
3
Certifications on fileISO 9001:2015, IATF 16949:2016, ISO 13485:2016, ISO 27001:2022.

The workflow

5 essential steps from CAD file to finished part

1
Step 1 – Design for powder removalModel internal channels with two open ends and a minimum diameter of 3–5 mm so loose powder can escape. Avoid blind pockets deeper than 3 times their width. Wall thickness should stay between 1 mm and 10 mm; thinner walls warp during sintering, thicker walls trap powder and shrink unevenly. Add a 0.5–1 mm machining allowance on every face that needs a tight final dimension.
2
Step 2 – Scale the model and orient the buildApply the shrink compensation factor from your alloy supplier, then verify it with a coupon on the first build. Orient the part so the longest axis lies flat in the powder bed. This reduces layer count and limits the effect of anisotropic shrinkage. Keep parts at least 3 mm apart so the recoater does not drag green edges.
3
Step 3 – Print, cure, and de-powderAfter printing, cure the whole build box at 180–200 °C for 2–6 hours to harden the binder. Let it cool, then dig the parts out of the powder cake. Brush off the bulk powder with soft brushes, then finish with compressed air at low pressure. Never use a metal scraper. Recover unused powder through a sieve for the next build.
4
Step 4 – Sinter under a controlled atmosphereRun the sintering cycle in vacuum or under argon or hydrogen, depending on the alloy. A typical profile ramps at 3–5 °C per minute, holds at 1,200–1,400 °C for 60–120 minutes, then cools slowly. A fast ramp cracks the part. A slow cool prevents oxidation. Check density after the run with Archimedes testing or a cut-up coupon.
5
Step 5 – Finish-machine the critical facesSintered parts go straight onto a 3-axis or 5-axis mill for datums, bores, threads, and sealing faces. Clamp on sacrificial stock rather than finished surfaces. Take light cuts first to confirm the shrink result before committing to final dimensions. This is where you recover ±0.005 mm tolerance and Ra 0.8–1.6 μm finish on the faces that matter.

Process choice

Binder jet vs CNC machining: which to pick

Use this table to route a part to the right process before you quote it.

Factor	Binder jet	CNC machining
Geometry	Internal channels, lattice, hollow shapes	Prismatic, turned, 5-axis contoured
Tolerance as-made	About ±0.5% of dimension	±0.005 mm
Surface finish	Ra 6–10 μm as-sintered	Ra 0.2–1.6 μm
Batch size	30–200+ parts per build	1 part to 10,000+
Material choice	Limited sinterable alloy list	Wide: aluminium, steel, titanium, plastics
Cost driver	Build volume and furnace cycle	Machine time and setup
Best when	Complex internal geometry at volume	Tight tolerance and finish required

FAQs

Frequently asked questions

What tolerance can binder jet metal 3D printing hold as-sintered?

As-sintered, expect roughly ±0.5% of the dimension. On a 100 mm feature that is about ±0.5 mm.

If a feature needs ±0.005 mm, it must be machined after sintering. Plan a 0.5–1 mm allowance on those faces.

Which alloys can be processed?

Stainless steels such as 316L and 17-4PH are the most common, along with tool steel and some nickel alloys.

Aluminium and titanium are possible but the sinterable powder range is narrower than what you can buy as bar stock. If your alloy is not on the sinterable list, CNC machining from solid is the safer route.

How do I stop parts from warping during sintering?

Keep wall thickness consistent, avoid isolated thin sections, and orient the longest axis flat in the build box.

A shrink coupon per batch tells you whether the furnace profile is stable. If a part still distorts, the design usually needs a rib or a thicker section, not a slower ramp.

Is binder jet cheaper than CNC for a single prototype?

No. A single part usually costs more on binder jet because the furnace cycle and setup are fixed.

Binder jet becomes economical when you fill a build plate with 30–200 parts. Below that, a 3-axis mill is normally the cheaper and faster path.

Can you machine a sintered blank that we print ourselves?

Yes. We finish-machine sintered blanks on 3-axis, 4-axis, and 5-axis centers, and we can hold ±0.005 mm on the machined features.

Send the blank dimensions, the alloy, and the drawing. We confirm the workholding and the machining allowance before quoting.