Datum
All skills
MaterialsRev. Q3 2026

Material selection tradeoff

Run a structured material selection tradeoff: translate the design problem into material indices, rank candidate metals and polymers against stiffness, strength, fatigue, corrosion, temperature, and cost, and state the failure mode each candidate is weakest against. Use whenever the user asks what material to use, whether aluminum can replace steel, if a plastic will survive, mentions weight reduction, or needs to justify a material choice to a reviewer.

Tomas Berg portraitTomas BergMaterials EngineerSandvik logoSandvik

Likes are peer vetting. Like this skill if it held up in your work.

Material Selection Tradeoff

Turns "what material should I use" into an Ashby-style comparison with the constraints written down, so the answer survives a design review.

Required inputs

  1. Function and loading: what the part does, dominant load type (tension, bending, torsion, impact, pressure), static or cyclic, and design life.
  2. Hard constraints: max temperature and exposure time, chemical/corrosive environment, regulatory or industry requirements (food contact, flammability, outgassing, weldability), and any fixed geometry.
  3. Objective: minimize mass, minimize cost, maximize stiffness in a fixed envelope, or a stated combination. If the user says "just make it good," ask which one they'd sacrifice first.
  4. Manufacturing route and volume: machined, cast, molded, welded, printed; 1 unit or 100k.
  5. Candidates already in mind, plus materials the company already stocks or has qualified. Incumbent materials get a seat at the table by default.

Method

  1. Select the material index for the geometry and objective. State it explicitly. Common cases:
    • Stiffness-limited beam in bending, minimize mass: E^(1/2)/rho
    • Strength-limited beam in bending, minimize mass: sigma_y^(2/3)/rho
    • Stiffness-limited tie in tension: E/rho
    • Strength-limited, minimize cost: sigma_y/(rho x cost per kg) If the geometry is a panel or shaft, use the corresponding exponent and say why.
  2. Screen on hard constraints first. Eliminate candidates that fail temperature, corrosion, or regulatory limits and record why. A material that melts is not a tradeoff.
  3. Rank survivors on the index using representative property values. Use typical values, state them, and flag properties that vary strongly with temper, grade, or fiber orientation. For polymers above roughly 50 C or under sustained load, use creep-adjusted modulus and flag it.
  4. Failure mode audit: for each finalist, state the failure mode it is most vulnerable to in this application (fatigue at welds for 6xxx aluminum, stress corrosion for high-strength 7xxx, notch sensitivity for hardened steels, UV embrittlement and creep for unfilled polymers, galvanic attack for carbon fiber against aluminum, hydrogen embrittlement for electroplated high-strength fasteners).
  5. Practical layer: availability in the required form (plate, bar, tube, resin grade), lead time risk, single-source flags, and processing consequences (heat treat distortion, weldability loss in T6, mold cost for a new polymer).
  6. Recommend one primary and one fallback, with the specific grade and temper or resin grade, not just the family: "6061-T6" not "aluminum," "30 percent glass-filled PA66, heat stabilized" not "nylon."

Output format

  • Constraint screen table: candidate, pass/fail per constraint, elimination reason
  • Ranking table: index value, relative mass or cost vs. incumbent, key caveat
  • One paragraph per finalist: weakest failure mode and what design detail mitigates it
  • Recommendation with grade/temper and the single test that would de-risk it

Guardrails

  • Never present handbook properties as guaranteed minimums. Distinguish typical vs. minimum values and recommend the material spec (ASTM, AMS) for anything structural.
  • If fatigue, fracture, or human safety governs, recommend design allowables from the applicable standard or test data, not typical values.
  • Composites and printed materials are anisotropic; refuse to rank them on a single scalar modulus without stating orientation assumptions.
  • If the honest answer is "the incumbent material is fine," say that.