You track the powder lot. You don't track the powder.

I walked into an additive facility last year that looked flawless on paper. Clean rooms. Documented procedures. Certificate of conformance for every powder canister in the inert storage cabinets. They could trace which supplier produced the atomisation heat, which ship delivered it, which operator unloaded the pallet. Immaculate traceability.

Then I asked how many times the powder currently sitting in that L-PBF machine had been cycled through a build.

Silence.

They didn't collect it. Not because someone forgot — because nobody had defined the parameter as worth tracking. This is the gap at the centre of aerospace additive manufacturing, and most of the industry is a few years behind the people who are already finding it during audits.

The certificate covers arrival, not survival

Here is the structural problem. When a lot of Ti-6Al-4V or Inconel 718 powder arrives at your facility, the certificate of conformance tells you what the atomiser measured at packaging. Particle size distribution. Apparent density. Hall flow rate. Chemistry, including interstitial oxygen and nitrogen. Real numbers, from accredited labs. They mean something — for the powder as it existed in a sealed container at the supplier's facility.

What happens next is where additive diverges from every other material process we audit. You sieve the powder. You load the build chamber. You run a build at 500 to over 1000 degrees Celsius in a controlled atmosphere that is never perfectly controlled. You recover powder from the build plate, sieve it again, return it to inventory. Repeat.

Each cycle changes the material. Oxygen content climbs as the powder surface reacts with residual oxygen and moisture in the chamber — and the activation energy for that oxidation is right there in the melt pool. Spatter particles coarsen the particle size distribution. Fines sinter and fragment. Flowability drifts, which changes how the recoater deposits each layer, which changes bulk density, which changes the thermal history of every voxel in the part.

The powder you qualified on arrival is not the powder forming your flight-critical bracket on cycle eleven. It is a different material. Different chemistry, different morphology, different rheological behaviour. And every one of those changes is invisible to the certificate you filed.

Your inspection sees the part. It can't see what the powder remembers

Post-build inspection is where most shops place their faith. For this problem, that faith is misplaced. CT scanning catches porosity and lack-of-fusion. Tensile testing catches gross mechanical deviation. Surface finish and dimensional metrology catch geometric nonconformance. None of these detect the full fingerprint of powder degradation.

If your material changed between qualification and production, the best inspection plan in the world is inspecting a part made from a material you never approved.

Oxygen pickup in titanium changes tensile strength and ductility in opposite directions — stronger and more brittle simultaneously. Both numbers can land inside the qualified envelope while fracture toughness has quietly shifted into territory your qualification programme never characterised. PSD drift changes fatigue crack initiation behaviour in ways that static tensile data will not reveal. You can pass every inspection on the router and still ship a part whose underlying material state diverged from qualification two or three sieve cycles ago.

In my PFMEA work on additive processes, the failure mode that scores highest on severity and detection is not the machine stopping mid-build. It is powder silently degrading across reuse cycles while every downstream control signals green. Detection ranking is high because no post-build test catches it. Severity is high because the failure manifests in fatigue life, not in a single static test. The standards already give you the language for this — what they do not give you is acceptance criteria for cycle twelve versus cycle three. You have to write those yourself.

Lifecycle control isn't hard. That's why skipping it is a choice

The systematic process control discipline that drove a 50% reduction in EASA audit findings in my own work rests on one principle: if a parameter affects product quality, you measure it, you control it, you hold the evidence. Powder lifecycle is not ambiguous. We know oxygen pickup matters. We know PSD drift matters. We know flowability correlates with layer uniformity, which correlates with porosity and surface roughness. The measurement methods exist. Oxygen and nitrogen analysis by inert gas fusion costs roughly €40 to €60 per sample at a commercial lab. Laser diffraction PSD analysis is €100 to €150. Hall flow is a stopwatch and a funnel — it costs nothing and takes under two minutes.

I have sat in rooms where the cost of that €40 sample was presented as a barrier. A single additive build plate of flight-critical parts represents five to six figures of value. The argument does not survive arithmetic.

What is actually missing is the decision to define powder as a process input that changes over time, and to set acceptance criteria at each sieve cycle. That is a management decision, not a technical one.

Aerospace additive is crossing from prototyping into flight-critical serial production. You can have perfect lot traceability, immaculate machine logs, and an inspection plan signed off by three people with impressive titles — and still ship parts made from a material you never characterised. Close the lifecycle gap, or the gap closes the programme.