1. Verification and Validation

In simulation governance, verification asks “did we solve the equations right?” while validation asks “did we solve the right equations?”. The former focuses purely on numerical accuracy, the latter on agreement with physical evidence.

Verification has two layers: code verification (vendor‐style regression and analytical benchmark tests that prove the software implements the governing equations correctly) and solution verification (user-run error estimation that quantifies the numerical uncertainty of a specific simulation).

By contrast, validation demands controlled comparisons of simulation outputs with high-quality experimental or field data, because every new geometry, load case or flow regime introduces fresh assumptions whose realism must be tested.

Although commercial vendors do substantial regression and code checks, industry standards such as ASME V&V 10/20/40 and NASA guidelines make clear that ultimate responsibility for additional verification, validation and uncertainty quantification lies with the analyst or organization deploying the model.

A disciplined V&V programme therefore builds quantitative credibility, lets teams expand confidently into multiphysics or high-consequence scenarios, and can justify reducing physical prototypes or accelerating certification cycles.