Titanium 3D Printing Guide: How the Process Works and Where It Stops

• 1
  Key takeaways
• 2
  What actually happens inside a titanium 3D printing build
• 3
  Design rules that keep a titanium build printable
• 4
  Heat treat, HIP and machining after the build
• 5
  Where titanium 3D printing stops making sense
• 6
  Choosing between printed and machined titanium

Both common titanium processes are powder bed fusion. A recoater spreads a thin layer of Ti-6Al-4V powder, a heat source melts a 2D slice of the part, and the build plate drops by one layer height. Repeat a few thousand times and you have a solid part. That is the whole idea, and nearly every limitation follows from it.

Selective laser melting uses a fiber laser, typically 200-400 W, in an argon chamber held below 0.1 percent oxygen. Layer heights run 30-60 um. Melt pools are small and cool fast, so the grain structure is fine and the as-built surface is relatively smooth for a metal powder process.

Electron beam melting uses a beam of electrons in vacuum instead. The bed is preheated to roughly 700-1,000 C, so the part stays hot through the build. Layer heights are coarser, around 50-70 um, and the surface comes out rougher. The payoff is much lower residual stress.

That temperature difference explains most process selection. If you need thin walls and fine channels, SLM is usually the answer. If you have thick sections that would crack or warp in a cold build, EBM handles them with less fight.

Overhangs are the first constraint. Anything steeper than about 45 degrees from the build plate needs support. Supports are printed in the same alloy and cut off later, which means every supported face needs finishing before it is a real surface. Place a part so the critical faces point up and the support scar lands on a non-critical face.

Wall thickness depends on the machine and the alloy. For Ti-6Al-4V on SLM, 0.4 mm is a practical floor for a stable wall, and 0.8-1.0 mm is safer when the wall carries load. Thin features also distort more, because the thermal gradient across a thin section is steeper.

Internal channels are the classic reason to print titanium. A 1.5-2 mm channel is printable if it can be cleared of trapped powder, and powder removal has to be designed in. A blind channel with no exit is a part you cannot clean. Add an escape path or accept that the channel stays packed.

Shrinkage is real but predictable. Ti-6Al-4V contracts roughly 0.5-1.0 percent from melt to room temperature, and machine software compensates with a global scale factor. Long thin parts still move unpredictably, so datum features should be machined after the build, not printed.

• 1
  Overhang under 45 degreesNeeds support; plan where the removal marks will land.
• 2
  Closed volumesAvoid them. Trapped powder cannot be removed and adds undocumented mass.
• 3
  Build orientationIt sets both the support layout and the anisotropy of the finished part.

A printed part is not a finished part. Stress relief comes first, usually a vacuum anneal around 800 C for two hours, which drops the residual stress left by rapid cooling. Skip it and a thin part will bow when you cut the build plate free.

Hot isostatic pressing is the next step for anything that sees cyclic load. HIP runs roughly 900-1,000 C at around 100 MPa of argon pressure for two to four hours. It collapses internal porosity below the resolution of a CT scan and brings fatigue life close to wrought material.

Then the part goes on a machine. Printed titanium is machinable but gummy: low thermal conductivity, high chemical reactivity, and a tendency to work-harden at the cut. Rough with carbide, finish with sharp edges and plenty of coolant, and expect tool life shorter than on 6061 aluminum.

At GreatLight we run the printed blank through our own 5-axis centers to hit the interfaces that matter. A build plate face is rarely flat enough for a mating surface. Machined datums, bores and seal faces bring the part to +/-0.005 mm while the printed geometry stays as-built.

If the part is a simple block with a few holes, printing is the wrong process. A block has no geometric complexity to exploit, so you pay for powder handling and post-processing without getting anything back. Machining from bar stock is faster and cheaper at any volume.

Size is the other wall. Powder bed machines have a fixed build envelope, and large titanium parts often cannot fit. GreatLight machines titanium up to 4,000 mm on the large travel, which covers parts no powder bed printer will hold. For anything long or wide, subtractive is the only route.

Volume economics flip one way for small parts too. At one to ten pieces, printing a bracket can beat tooling costs. Past a few thousand identical parts, casting or machining usually wins on unit price, because print time per part does not drop with quantity.

There is also inspection cost. A printed titanium part often needs CT scanning to verify internal channels or porosity, and that adds time and money that a machined part never sees. If your quality plan cannot absorb CT, keep the design machinable.