Ultra-low porosity stainless steel and UniJet SS17-4PH: what changes on the shop floor

• 1
  Where porosity comes from in ultra-low porosity stainless steel
• 2
  What density numbers do and do not tell you
• 3
  Machining printed 17-4PH without making porosity worse
• 4
  When a low-porosity powder is not the answer
• 5
  Surface finishing and inspection for porous stainless parts
• 6
  Print or machine: matching the route to the feature

Porosity in a printed stainless part is trapped void space. It starts as gas or as unfilled gaps between powder particles. Laser powder bed fusion melts a thin layer of powder, and if the melt pool does not fully wet the particles below it, small cavities survive. If the powder carries internal gas from the atomization step, that gas expands in the melt and leaves a spherical pore behind.

UniJet SS17-4PH is a gas-atomized 17-4PH (SUS630) powder. The ultra-low porosity claim is mostly about two things: particle size distribution and the fraction of hollow or gas-filled particles. A tight distribution packs better, so the laser sees a denser bed. Fewer gas-bearing particles means fewer pores that no scan strategy can close.

17-4PH is a martensitic precipitation-hardening grade. Copper precipitates form during aging and raise strength without a full quench. That makes it popular for brackets, housings, and tooling inserts. It also means the alloy behaves well under a laser, because the melt pool is stable across a wide parameter window.

For a machinist, the practical question is not the marketing density number. It is whether the part in front of you has open surface pores, subsurface voids, or neither. Those three cases need different handling on the mill.

• 1
  Gas porositySpherical voids from trapped argon or from hollow feedstock particles.
• 2
  Lack-of-fusionIrregular voids between layers when energy density is too low.
• 3
  Keyhole porosityDeep voids from excessive energy density and unstable melt flow.

Density is reported as a percentage of the wrought value, usually by Archimedes method. A reading of 99.5% sounds close to perfect. On a 100 mm part, that 0.5% is still real void volume, and it does not tell you where the voids sit. A cluster of pores near a fillet is a fatigue problem. The same volume spread evenly through the core is often harmless.

Archimedes testing also struggles with open surface pores. If coolant or water enters a surface pore during measurement, the result skews. That is one reason two labs can report different numbers for the same build.

The useful companion number is pore size and location from CT scanning. Micro-CT at 5–10 μm resolution shows whether pores are isolated or connected. Connected porosity is the one that leaks, and it is the one that ruins a pressure-tight part.

So when a supplier quotes a density figure, ask three follow-ups: measurement method, sample size, and whether the part was HIPed. Hot isostatic pressing closes internal voids but does nothing for a pore that already broke the surface.

Cutting a printed part is not the same as cutting a forging. The as-built surface is rough, often Ra 8–12 μm, and it can hide partially fused particles. A first pass that skims 0.2 mm may just smear those particles into the surface. Take a real cleanup cut, 0.3–0.5 mm, and let the tool get under the crust.

17-4PH in the as-built condition machines like a soft stainless, roughly 30–36 HRC. In the aged condition it moves to 40–44 HRC depending on the aging cycle. Tool life and surface finish change a lot between those two states. Decide the heat-treat sequence before you write the program.

Coolant matters more than usual. If the part has open surface pores, flood coolant pushes fluid inside, and it comes out later as a stain or a corrosion site. For porous parts, consider dry or minimum-quantity lubrication on finishing passes, then clean in an ultrasonic bath.

Thin printed sections can also move when the crust is removed. The stress state in a printed wall is not symmetric, so relieving one side releases the other. Rough, then semi-finish, then finish, with a stress-relief step between if the wall is under 1.5 mm.

• 1
  Stock allowanceLeave 0.5–1.0 mm on printed faces that will be machined.
• 2
  ClampingSupport thin walls from both sides; printed ribs crush easily.
• 3
  Heat treat firstAging before finishing avoids distortion of a finished surface.

Low porosity helps fatigue life, pressure tightness, and polished surfaces. It does not fix a bad build orientation, and it does not remove the need for support removal. If the part geometry already forces a rough downskin, a better powder only moves the finish a little.

For a simple bracket with three holes and a milled face, printing is the wrong route regardless of powder quality. Bar stock 17-4PH at Ø20–80 mm is widely available, and a 3-axis mill holds the drawing in one setup. Print time, support removal, and finishing add cost with no benefit.

There is also a size ceiling. Our largest machining travel is 4,000 × 400 × 150 mm, and 5-axis work covers most brackets and housings. Printing a part that size in stainless is slow and distortion-prone, so the hybrid route, print a near-net blank then machine the critical faces, is usually the practical one.

The honest test: list the features that must be tight, the features that must be hollow, and the annual quantity. If the tight features dominate, machine. If the hollow features dominate and the tight faces are few, print then machine those faces.

If a printed surface must be sealed, the sequence is: remove supports, stress relieve, machine or grind the functional faces, then seal or plate. Plating a porous surface traps chemistry in the pores, and it bleeds out over weeks. Ultrasonic cleaning before plating is not optional.

For cosmetic parts, bead blasting followed by tumbling gives a uniform matte that hides small surface pores. Polishing to a mirror finish does the opposite. It opens pores and makes them visible, so a mirror finish on a printed stainless part should be quoted as a risk, not a routine operation.

Inspection should match the risk. A dimensional report covers the drawing. A leak test covers pressure tightness. CT covers internal voids. Sending a CT scan on every part is wasteful; sending it on the first article of a new geometry is cheap insurance.

We run 100% inspection before shipment, with raw material checks, in-process monitoring, and a final pass. Reports are available on request. For 17-4PH we can also handle the aging cycle and document hardness.