agentleFS
BaratiLabBaratiLab/mechanical-engineering-skills/skills/am-alloy-screening/SKILL.md

am-alloy-screening

Skill16 starsChanged 16 days agoMITView on GitHub

Screen metal alloys — known grades or new compositions — for laser powder bed fusion printability. Parse a composition, get thermophysical properties (Thermo-Calc/TC-Python CALPHAD when available, built-in alloy data or flagged estimates otherwise), compute Rosenthal melt pools and lack-of-fusion process maps over power and speed, and rank candidates by process window with explicit, unit-checked outputs. Based on the Barati Farimani Lab's agentic AM alloy evaluation. Use for "which alloy can I print for X", new-composition feasibility, LoF process maps, or checking material property files.

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Pin down the candidates: If the user names an application (corrosion-resistant, high-temperature, lightweight), propose 2–5 candidate …

    line 19

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

AM alloy screening

Pak et al. (Additive Manufacturing Letters 17 (2026) 100355) let an LLM agent chain these steps for any alloy, known or hypothetical:

  1. Thermo-Calc thermophysical properties
  2. a Rosenthal melt-pool model
  3. a lack-of-fusion (LoF) process map

You are that agent. scripts/alloy.py provides the deterministic tools, and it needs only numpy and matplotlib. Paper details, plus a unit bug found in the paper's own outputs, are in references/paper-notes.md.

Workflow

  1. Pin down the candidates. If the user names an application (corrosion-resistant, high-temperature, lightweight), propose 2–5 candidate alloys and say why each fits.
    • known lists the built-in grades: SS316L, 17-4PH, H13, IN718, IN625, Hastelloy X, Monel K500, Ti-6Al-4V, AlSi10Mg, Al 7050, CuSn10, Cu, W, CoCrMo.
    • New compositions: compose "Fe-10Ni-10Ti" (balance element first, wt%), compose Al90Cu5Ti5, or add --atomic for at%.
  2. Get properties.
    python scripts/alloy.py properties --alloy IN625 --out IN625.json
    python scripts/alloy.py properties --comp "Fe-10Ni-10Ti" [--thermocalc] --out FeNiTi.json
    • With Thermo-Calc and TC-Python (references/thermocalc.md), the database is chosen by the paper's rules. You get solidus and liquidus from the liquid fraction, density, Cp, k, and absorptivity via Drude from resistivity.
    • Without it, density and Cp come from the rule of mixtures; k, solidus, liquidus and absorptivity come from the nearest built-in alloy, with a confidence flag. Pass a low-confidence warning on to the user.
  3. Check units: check mat.json (it also runs automatically). It catches density in g/m³ and Cp in J/mol·K, the slips found in the paper's repository outputs.
  4. Map and rank.
    python scripts/alloy.py map  --material IN625.json --hatch 50 --layer 30 --out in625.png
    python scripts/alloy.py rank --materials IN625.json SS316L.json --hatch 50 --layer 30 --out compare.png
    • The defaults are the paper's: P and V from 100 to 1000 (W, mm/s) in steps of 100, at the layer and ±25 µm.
    • Widen the ranges (--power 50:500:25 --velocity 200:2000:100) when every alloy has an empty LoF region, otherwise the ranking only reflects tie-breakers.
    • rank sorts by the LoF fraction at the prescribed layer, then over all layers, then by the mean overlap criterion.
  5. Recommend from the keyed numbers, not from impressions of the plot.
    • A smaller lack_of_fusion_fraction (a larger process_window_fraction) is better. The paper's one wrong recommendation came from misreading this.
    • Give the property source and confidence, and suggest single-track and density-cube trials before qualification.

Limits (state them)

  • Rosenthal is conduction mode only. The LoF boundary is conservative, and keyholing and balling are not predicted. For those, pair this skill with lpbf-defect-prediction and meltpool-geometry.
  • Absorptivity matters a lot for Cu and Al at IR wavelengths. The Drude estimate is a baseline, and real absorptivity rises sharply once keyholing starts.
  • New compositions need CALPHAD or measurement. The fallback says so, and you should too.

Cite: Pak, Chandrasekhar & Barati Farimani, Addit. Manuf. Lett. 17 (2026) 100355.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.