JONA ENGEL. Metallurgy note Profile
Laser metallurgy triage

When the process goes sideways, decide what to prove next.

A compact triage surface for laser cladding and DED: material plus machine settings become a process condition, a chemical risk, a structure/property hypothesis, and a decision path for this week.

For CEOs, members of advisory boards, commercial SVPs, and CTOs under pressure: the useful question is rarely "who is right?" It is which failure hypotheses are plausible, which evidence can separate them quickly, and what decision becomes safe enough to make.

Signal

Cracks, poor wetting, inconsistent build height, supplier disagreement, or a customer deadline with weak evidence.

Hypotheses

Thermal window, dilution, shielding, chemistry, microstructure, surface state, or simply the wrong material/process route.

Evidence

Cross-section, chemistry, hardness map, crack check, supplier records, and a narrow coupon matrix.

Decision

Continue, pause, change route, ask the supplier a sharper question, or escalate to a real lab proof.

01 · Who lands here

Different roles, same problem: uncertainty is blocking action.

The same problem has to be legible to a boardroom, an operations team, and the specialist who will own the proof. The model is deliberately reduced: it keeps the decision logic visible and leaves expensive precision to measured projects.

CEO / Advisory Board

Need decision relief

They want a clear read on whether a process problem is containable, whether the available explanation is plausible, and what proof reduces the next decision risk.

CTO / Ops

Need the failure tree

They need the shortest path from symptom to likely mechanism: thermal input, chemistry, HAZ, microstructure, feed, shielding, or supplier process drift.

Founder / Supplier

Need credible choices

They need to choose material, route, and test plan before burning weeks on a promising sample that later fails under the buyer's real constraints.

02 · Build a track

A compact decision model for laser cladding / DED.

The sketch separates two views that are easy to confuse. First it shows the fast scan direction: the source moves, a molten pool trails, and the raised track accumulates behind it. Then it pivots to the slow-axis section where adjacent passes overlap, remelt older flanks, and leave a scalloped top, fusion line, dilution zone, and HAZ band. The top view below carries the literal scan path and HAZ wake. It is a schematic of mechanism and direction: no hidden calibration, no exact temperatures, no claim that the model replaces a section and a microscope.

Laser cladding track and overlap story A schematic laser-cladding / DED scroll story: a fast-scan side view where the moving source builds one track, a restrained section pivot, and a slow-axis cross-section with overlap, dilution, heat-affected zone, and curved remelt interfaces between passes. track build → overlap section decision output substrate original surface · datum fast scan · track builds behind source focus laser scan direction · bead cools behind source section view rotates across track spacing fast scan slow-axis section slow-axis overlap · remelt interfaces fusion line / dilution HAZ band track index direction · overlap sets remelt 900 W · 1.0 m/min · 2.2 mm spot · 7 g/min material: IN713 hard mode process: conduction shielding: controlled this week's decision Pause claims red-team the route watch first Liquation / hot crack surface history proof step Crack check first then chemistry uncertainty: high route high-risk superalloy
hot
fusion window
balanced
bead build
medium
HAZ / remelt
high
risk state
High-risk window. Do not turn this into a customer promise. Separate thermal geometry from chemistry and cracking with a deliberately small proof plan.
Top view: tracks, melt pool and HAZ trail A top-down view of the deposited tracks, the elongated melt pool at the laser, and the trailing heat-affected-zone isotherms that stretch downstream as scan speed rises. top view · melt pool & HAZ trail scan → melt pool HAZ trail

Build mode

cladding · 1 layer

Material family

IN713 hard mode

Laser power

900 W
300 Wnominal3 kW

Scan speed

1.0 m/min
0.3nominal3.0 m/min

Focus / spot

2.2 mm

Powder / filler feed

7.0 g/min

Track overlap

35%
gapsremelt

Shielding

controlled

Model discipline: these are dimensionless, binned estimates from simple energy, mass-per-length, overlap, and material-family rules. A real process window needs measured absorptivity, section geometry, powder catch or filler delivery, shielding records, chemistry, microscopy, and property targets.

03 · What gets covered without clutter
Machine -> condition

Geometry first

Power, speed, spot, feed, and overlap become fusion, wetting, build height, dilution, remelt, and HAZ. That geometry is the first filter before chemistry and property risk.

Condition -> chemistry

The hidden clock

O/N uptake, Cr-depletion (sensitisation), segregation, and carbide reactions set up the real failure modes — hot/solidification and liquation cracking, cold/hydrogen cracking in hardenable steels, and thermal-shock / residual-stress cracking. These ride on chemistry and thermal history (preheat, restraint, cooling rate), not bead shape, and can dominate even when the bead looks acceptable.

Chemistry -> risk

Properties decide

The useful output is a testable property hypothesis: crack risk, ductility loss, corrosion risk, hardness scatter, wear behavior, or fatigue uncertainty.

Risk -> evidence

Next proof

The model ends by naming the measurement: cross-section, EDX/LECO chemistry, hardness map, crack check, supplier process records, or a smaller coupon matrix.

Evidence -> decision

Commercial action

The point is not prettier simulation. It is deciding whether to continue, pause, switch material, change process route, or ask a supplier a sharper question.

Uncertainty visible

No fake certainty

If the missing variable is shielding, absorptivity, powder catch, composition, or surface preparation, the page says so instead of laundering it into a false number.

04 · Honest boundary

Useful for triage, not a substitute for a qualified process window.

This model can do

Make the mechanism map visible, show why material choice changes the risk, explain why chemistry can beat geometry, and help a non-specialist understand the proof stack before a meeting.

This model cannot do

Certify a weld, predict exact peak temperature, replace FEM/CFD, validate IN713 weldability, guarantee corrosion or fatigue performance, or turn unknown shielding and chemistry into reliable advice.

Material-specific cues: IN713 is a high-risk gamma-prime nickel superalloy where liquation, hot cracking, segregation, and strain-age cracking deserve early proof. Ti-6Al-4V is shielding and oxygen/nitrogen sensitive. Stainless routes need passivity/sensitisation (Cr-depletion) and grade context. Pure copper is a coupling problem first: low near-infrared absorptivity and high thermal conductivity push lack-of-fusion at low energy, then porosity and spatter risk when power density is forced high. The honest proof is a section + crack check, hardness and hydrogen for hardenable steels, ferrite number for stainless, copper density/conductivity and coupling checks, and wavelength, preheat, surface, and shielding records.

05 · Contact

Questions, corrections, or public-source pointers are welcome.

Especially useful: examples where a section, chemistry check, hardness map, or supplier record changed the interpretation of a laser-cladding / DED problem.

Privacy & contact note