material-substitute

AI skill that suggests material substitutions when the specified material is unavailable, on long lead time, too expensive, or unsuitable. Returns 2–4 candidates with property comparisons (yield, tensile, hardness, corrosion class), machining notes, relative cost, lead time, and explicit customer-approval implications (ECN, MRB disposition, partial FAI requirements for aerospace work).

Use case

Tuesday afternoon. The PO calls for 17-4 PH H1025. The supplier is on backorder until next month. The customer's PO is firm in 3 weeks.

The shop owner has three options: tell the customer "no", switch to a substitute, or pay 4× for a small-lot specialty supplier. Picking a substitute is a 30-minute decision involving the buyer, the engineer, and (for aerospace) the customer's source-substitution rules. This skill collapses the first 25 minutes — what's actually substitutable, what does it cost in properties, and what approvals are needed before swapping.

Critical: the skill is honest about tradeoffs (no "equivalents", just better/worse on specific dimensions) and refuses to recommend substitution for aerospace or medical work without flagging the approval requirements. For some materials (Inconel for high-temp, titanium for biocompatibility), it will say "do not substitute; extend lead time."

Invocation

• Messaging: @material-substitute 17-4 PH H1025, stockout, 3-week lead time, aerospace bracket
• With context: @material-substitute customer specs 304 SS but stockout, part is a non-pressure indoor enclosure, qty 50, lead-time critical
• REST: POST /api/skill/run

Minimum useful input

• Original material with condition (17-4 PH H1025, 6061-T6, 304 SS annealed)
• Reason for needing a substitute (stockout, cost, lead time, machinability)
• What the part does (structural, corrosion, wear, thermal, electrical, biocompatible)
• Industry / approval class (aerospace, medical, automotive, commercial)

Without "what the part does," the skill can give a generic list of property-equivalents but can't recommend one — the function matters more than the spec.

Sample output

User: @material-substitute 17-4 PH H1025 bracket for aerospace fitting, stockout, 3-week lead. Yield ~170 ksi needed, mild corrosion exposure.

Skill:

Original material: 17-4 PH SS, condition H1025
  Yield: 170 ksi, Tensile: 185 ksi, Hardness: 38-42 HRC
  Corrosion: moderate (similar to 304 — no chlorides)
  Common in: aerospace fittings, valve bodies, structural brackets

Substitution candidates

1. 15-5 PH SS, condition H1025
   Yield: 165 ksi, Tensile: 180 ksi, Hardness: 36-40 HRC
   Why: nearly identical PH steel family, slightly tougher transverse
   properties, sometimes more available than 17-4
   Cost: 1.05-1.15× of 17-4
   Lead time: typically similar; check your AVL
   Machining: identical to 17-4

2. 13-8 Mo PH SS, condition H1000
   Yield: 195 ksi, Tensile: 210 ksi, Hardness: 42-46 HRC
   Why: higher strength than 17-4, similar corrosion, often used
   interchangeably for "more capability than 17-4" applications
   Cost: 1.3-1.5× of 17-4
   Lead time: similar to 17-4 availability
   Machining: slightly harder, expect 10-20% lower MRR

3. 455 PH SS, condition H950 or H1000
   Yield: 200 ksi+, harder, less corrosion than 17-4 in chlorides
   Why: drop-in for higher strength but corrosion is lower
   Cost: 1.3× of 17-4
   Machining: noticeably harder; tool life lower

Customer-approval risk (aerospace):
  Substituting 17-4 with 15-5 or 13-8 is COMMON in aerospace but
  requires:
    - Customer (prime) written approval for the source substitution
    - Engineering verification: yield + tensile + corrosion meet all
      drawing callouts
    - Partial FAI per AS9102 against the substitute (Form 2 line
      changes, Form 3 mechanical retest if drawing calls for tensile)
    - Updated cert package (mill cert for the substitute material)
  Don't ship until you have written approval from the customer's QE.

Tradeoff summary:
  15-5 PH H1025 is the closest match by properties — if you have it
  available, swap with minimal risk. 13-8 buys you more capability at
  ~1.4× cost and may be overkill. 455 trades corrosion for hardness;
  if the bracket sees moisture exposure, not the right move.

Recommendation:
  15-5 PH H1025 if available. Get written customer approval before
  ordering. Add a partial-FAI line to the QA package noting the
  substitution. If 15-5 isn't available either, extend lead time
  rather than dropping below the PH-steel family — substitutes from
  300-series or 400-series won't carry the same fatigue performance.

Strengths and weaknesses

Customization

• AVL integration — if your shop has an approved vendor list with specific suppliers per material, add it to SKILL.md so the skill recommends from your approved sources.
• House cost data — replace the rules-of-thumb cost ratios with your actual buyer's recent pricing.
• Customer-specific source rules — Boeing D1-4426 has its own substitution rules; if your shop quotes a lot of Boeing, add a section.
• Composite / plastic / ceramic depth — current SKILL.md is strongest on metals. If your shop runs more plastics or composites, extend the body with the families you see often.

Pairs with

• quote-estimator — when the customer's material is unavailable, this skill suggests subs; the estimator re-quotes with the alternative cost
• spec-summary — for the AMS spec the substitute material needs to meet
• gdt-explainer — if a substitution changes which features are critical
• rfq-triage — extracts the material callout from incoming RFQs; this skill is the next step when sourcing flags a problem

Source

See SKILL.md.