Data Governance & Verification Methodology
Engineering materials data directly impacts human safety, structural reliability, and capital investment. Mat Grades treats materials data governance as an uncompromising engineering standard.
1. The Principle of Non-Fabrication
If an ultimate tensile strength, electrical conductivity rating, or thermal expansion coefficient is unverified in primary technical references, Mat Grades leaves the record blank or keeps the material unverified. Under no circumstances are numerical properties interpolated, estimated, or synthesized using generative AI models.
2. Source Hierarchy & Authoritative-Body Preference
When compiling engineering data, Mat Grades enforces strict priority for authoritative bodies over secondary web sources:
3. Verification Lifecycle & Publication Integrity Gate
Mat Grades enforces a deterministic integrity architecture established in Phase 2B.2A. A material candidate cannot transition to PUBLISHED and VERIFIED until 100% of its engineering facts (chemistry, mechanical properties, physical constants, designations, standard specifications) possess valid item-level source citations.
Audited against primary authoritative literature (e.g. CDA application data sheets). Every property row links directly to an authoritative source record. Published to public index.
Staged candidate records or internal architecture tests. Strictly excluded from search engines (robots noindex), public catalog, and comparison selectors until verified.
4. Metallurgical Context Preservation (Non-Isotropic Data)
A single alloy designation (such as C11000 ETP Copper or C17200 Beryllium Copper) does not possess a single universal yield strength or tensile ductility. Mechanical properties are fundamentally governed by:
- Heat Treatment & Temper: (e.g. O60 Soft Annealed, H02 Half Hard, H04 Hard, TF00 / TH04 Precipitation Heat Treated)
- Product Form: (Flat products/strip, Rod, Bar, Wire, Tube, Extrusions, Forgings)
- Section Size & Thickness Range: (Due to work hardening gradients and grain size variations)
- Test Methodology: (ASTM E8 tensile test offset yield 0.2% vs extension under load 0.5%)
Mat Grades preserves contextual groupings so engineers can evaluate exact product forms rather than fictitious blended averages.
5. Discrete Hardness Scale Governance
All hardness values are published strictly on their cited measured scale (e.g. 85 HRF, 40 HRB, 78 HR30T). Cross-scale conversions must only be made using published ASTM E140 and ISO 18265 empirical tables.
6. International Equivalence & Cross-Reference Taxonomy
To prevent dangerous substitution errors in manufacturing, Mat Grades classifies international cross-references into four distinct categories: