Boundary-Condition Fidelity and Reynolds-Number Effects in CFD Validation for Vehicle Aerodynamics

Abstract

Credible validation of computational fluid dynamics (CFD) in vehicle aerodynamics depends on the engineering decision being supported. For a fixed vehicle at one operating condition, total drag with quantified uncertainty may be sufficient when only that force is required. It is insufficient, however, for explaining a flow mechanism, predicting a design delta, ranking closely spaced concepts, or transferring a model across geometries and conditions. Starting from the nondimensional governing equations, this paper examines three linked causes: geometric scaling changes Reynolds number and relative boundary-layer thickness; ground, wheel, and tunnel treatments change the represented physical system; and local errors may cancel in an integrated force. Evidence from Ahmed-body, detailed-vehicle, wind-tunnel, moving-ground, and rotating-wheel studies is evaluated through comparison. A decision-oriented framework then separates numerical verification from physical validation and links the required evidence to the intended use. Vehicle CFD is best treated as a conditional prediction whose credibility is bounded by geometry, operating state, physical setup, quantity of interest, and decision.