Commercial view
Budget needs evidence
Is the process route plausible at the volumes in question? Which proof would make the next commitment more defensible?
Metallurgy notes
How I think through metallurgy decisions.
Public notes on process routes, evidence, uncertainty, and the translation layer between metallurgical physics and market reality. The aim is to make the reasoning legible enough that the next proof becomes sharper.
Signal
Volume crossover unclear, supplier qualification stalled, repair vs replace economics weak, or a new alloy/process route with no defensible priority.
Hypotheses
Wrong process for the volume band, chemistry or shielding gate invisible to the model, powder supply concentration risk, or the commercial story does not match the technical one.
Evidence
Public sources, TCO breakpoints, published qualification data, visible adoption constraints, and a narrow proof matrix.
Next proof
Which narrow observation, section, source, or conversation would change the decision weight instead of adding another decorative slide.
01 · Who this is for
People who need the technical story to survive contact with cost, qualification, and adoption.
Engineers, founders, operators, researchers, and commercially minded technical people in advanced materials, surfaces, additive manufacturing, and industrial equipment. Long cycles, high stakes, and the uncomfortable truth that most opportunities do not work.
Technical view
The failure tree matters
Why is qualification stuck? Is the bottleneck geometry, chemistry, shielding, or the wrong process for the real production volume and post-processing reality?
Research view
Precision needs humility
What can be inferred from public data, what needs a section or experiment, and where does the model stop being honest?
02 · Reading map
Three lenses: route choice, confidence, and proof.
Read this as a compact map for how I separate process economics, metallurgical mechanisms, and the proof that would move a real decision.
Route choice
Process selection is capital allocation
PBF, DED, Binder Jetting, and repair/hybrid routes solve different problems. The right comparison is volume, geometry, post-processing, qualification, and supply reality.
Confidence
Not every public fact carries the same weight
A useful knowledge map keeps provenance, confidence, and what would change the confidence. Otherwise it becomes content, not evidence.
Proof
The smallest useful proof beats broad reassurance
A cross-section, chemistry check, hardness map, crack check, or supplier record can be more useful than another strategic narrative if it separates the live hypotheses.
Working model
Laser cladding / DED as the first grammar
The companion page turns machine settings and material family into binned geometry, process condition, risk, uncertainty, and a next-proof action. Open the DED/cladding grammar.
Boundary
Public note, not advice
Ranges are directional and public-source based. Any real capital, supplier, or qualification decision still needs primary data, measured sections, and current source checks.
Contact posture
Corrections are welcome
Questions, corrections, and public-source pointers are welcome through the contact note linked in the footer.
03 · Directional signals (2026 synthesis)
Metal AM and near-net-shape processes are moving from low-volume/high-value into the hard work of scale, TCO, and supply resilience.
Synthesized from public reports (Wohlers 2025, AMPOWER-linked data, VoxelMatters, IDTechEx and related 2025/2026 analyses). Ranges, not point forecasts. Always triangulate with the latest primary data before capital decisions.
Market size (2025/26)
USD 5–17B+ range
Hardware + materials + services/parts value chain. Narrower hardware+materials views sit lower; full value chain higher. Temporary softness in some hardware segments in 2024, but material consumption growth remains double-digit.
Growth outlook to ~2035
15–25% CAGR (directional)
Consensus points to continued strong expansion as throughput scales (multi-laser PBF, larger BJ systems, advanced DED), post-processing costs fall, and qualification matures. Asia-Pacific (esp. China) is a major installed-base driver.
Material consumption
~1,000–1,500+ tonnes / yr
Up from ~600 t in 2022. High-value alloys (Ti, Ni superalloys) dominate current CHF value; high-throughput routes (BJ, advanced DED) will shift the mix toward mid-value alloys while lifting total volume and value.
Laser cladding / surface-focused DED participates in a parallel surface engineering & repair market (~USD 0.6–0.7B in 2025, ~10% CAGR). High-growth themes: throughput scaling, hybrid manufacturing & repair economics, TCO (especially post-processing), functional surfaces / graded deposits, sustainability (recycled powder, buy-to-fly reduction, life extension), and supply-chain diversification for critical alloys.
04 · Technology triage — when to use what
A practical map for volume, TCO, and risk. Not a substitute for your own qualification and supplier data.
typical part value ↑
ones–hundreds / yr · high value
Laser Cladding / DED + Repair/Hybrid
One-off and small-batch economics on high-value components: repair, life extension, buy-to-fly reduction, large formats, multi-material/graded surfaces, adding features to existing parts. Repair/hybrid (DED + machining) is a major economic lever for turbines, molds, tooling. High deposition rates; lower resolution, more post-machining.
Use when: part size, repair economics, or surface functionality (wear/corrosion/thermal) are the value drivers. Strong hybrid play with machining.
hundreds–thousands / yr · high value
Powder Bed Fusion (PBF / LPBF / EBM)
Best for complex, high-value, certification-heavy parts in low-to-medium volumes (roughly 1–5,000 parts/year per design/batch; multi-laser pushing toward medium serial — this corner of the map is contested). Highest precision and properties; supports lattices and topology. Limitations: cost, speed for simple/high-volume parts, supports and thermal management.
Use when: geometry complexity or certification path dominates, and volumes fit the sweet spot. Uncertainty on exact medium-volume crossover remains medium today.
ones–tens / yr · lower value
Metal FFF / Bound Metal Deposition
Low-volume, prototyping, tooling, jigs, education/entry industrial. Low barrier, rapid iteration; final density and properties below the fusion routes.
Use when: accessibility and iteration speed matter more than final density.
thousands–100k+ / yr · mid value
Binder Jetting (Metal BJ)
Positioned for medium-to-higher volumes (thousands to tens/hundreds of thousands/year) where MIM-like economics can apply. No supports in build; strong cost-reduction potential at scale. Limitations: shrinkage/distortion in sintering, consistent full density (HIP often required), post-processing intensity.
Use when: simpler geometries and higher volumes are the target, and sintering/HIP/qualification path is viable. Highest volume-scalability potential among core technologies to 2035.
typical parts per year →
Quadrant placements are typical, not boundaries — real positions depend on alloy, geometry, and qualification path; the high-value × high-volume corner is where multi-laser PBF and BJ+HIP are both pushing. Post-processing (sintering, HIP, machining, heat treatment) frequently represents 30–70% of total cost and lead time across routes. The useful output of any triage is a defensible 90-day proof plan, not a simulation.
05 · Supply & process knowledge map
A small public slice showing confidence bands, not a complete decision record.
HIGH confidence (established)
Tungsten produced by hydrogen reduction — WO₃ + 3H₂ → W + 3H₂O is standard industrial powder-metallurgy practice. Not a new variable for most users of refractory PM.
MEDIUM → HIGH post-2028
Binder Jetting reaches medium-volume serial auto/industrial — volume scalability is real once sintering distortion, HIP consistency, and qualification cycles are solved at the specific alloy + geometry. Medium confidence in 2026 data; high once more production data exists.
HIGH (China dominance)
HREEs (Nd, Pr, Dy, Tb) sourced from ion-adsorption clays (China-dominant for heavy REEs). Separation remains the bottleneck. Recycling rates growing but still small relative to demand growth from EVs and wind.
HIGH susceptibility (specific grades)
Ti-6Al-4V susceptible to hydrogen embrittlement via hydride formation — especially in β-rich conditions or sustained load. Mechanism and diffusivity data are solid; component-level risk depends heavily on environment, load, and surface state.
MEDIUM (policy + execution)
H₂-DRI + EAF with green H₂ has an LCA impact of significant CO2 reduction (~90% vs BF-BOF). HYBRIT-scale plants targeting 2026+; electricity cost and electrolyzer scale are the practical variables.
HIGH (mature industry)
PGMs recycled by high-efficiency pyro/hydro — ~30%+ global supply from recycling, up to 95% recovery in autocatalysts. Pyromet or hydromet dominant. Strong circular model already operating at scale.
This is a public slice of a local knowledge base. The important discipline is not the graph itself; it is keeping source, confidence, and uncertainty visible so a claim does not quietly gain more authority than it deserves.
06 · Proof of method — first deep dive
Laser cladding / DED triage model (interactive grammar).
Material + machine route translated into binned fusion/build/HAZ state, chemistry and cracking hypotheses, named proof step, and explicit uncertainty. No fake temperatures or pool depths. The same decision grammar (symptom → hypotheses → evidence → this-week decision) that the market work rests on.
See the working model
Open the live laser-cladding / DED triage surface. Power, speed, spot, feed, overlap, shielding, and material family (IN713, Ti-6Al-4V, 316L, pure copper) drive geometry, regime (conduction/transition/keyhole), and risk state. Outputs are deliberately binned with named next-proof actions.
Open the deep-dive modelWhat it demonstrates
The public model separates thermal geometry from chemistry and cracking risk, states what cannot be known (absorptivity, powder catch, shielding records, final properties), and ends with a practical proof plan rather than a pretty picture. Executives get decision relief; technical teams get the shortest credible evidence path.
This is the reusable demo grammar. Later deep dives (other processes, alloy families, supply nodes) will follow the same honesty discipline.
07 · Method boundaries
What this public note can and cannot do.
The page is deliberately useful but incomplete. It can show the reasoning grammar, make uncertainty visible, and point to plausible proof steps. It cannot replace measured sections, supplier records, current primary sources, or a real qualification plan.
What is public
Directional synthesis
Public reports, technical literature, and visible process economics can narrow the question, but they should stay in ranges rather than pretend to be exact.
What is missing
Primary evidence
The decisive facts are usually specific: actual section geometry, chemistry, shielding records, supplier capability, customer constraints, and failure history.
What matters next
A proof that changes weight
The useful next step is the smallest observation or experiment that separates competing hypotheses and makes the decision less theatrical.
Questions, corrections, and public-source pointers are welcome. If a source looks stale, a term is off, or a claim carries too much confidence, point me to the better reference.
08 · Source discipline
How to read the confidence bands.
Market reports and technical papers are useful only when their provenance, date, uncertainty, and decision weight stay visible. The point is not to collect facts; it is to know how much weight each fact can carry.
Provenance
Date the claim
Adoption curves, capacity announcements, powder prices, and qualification claims decay quickly. A useful note keeps the source trail close to the conclusion.
Confidence
Keep ranges visible
When the evidence is directional, the language should stay directional. False precision is worse than an honest band.
Revision
Let better evidence move the map
A knowledge graph is only useful if new measurements, better sources, and failed assumptions can change the weight of a node.
09 · Contact
Questions, corrections, or public-source pointers are welcome.
Especially useful: better primary sources, cleaner terminology, or examples where a small proof changed a materials decision.