Vehicle cabins are key microenvironments for airborne disease transmission, owing to their confined spaces and high occupant density. Ventilation that reduces infection risk often causes drafts, creating a dilemma between health protection and thermal comfort. This study addresses the trade-off between airborne cross-infection risk and thermal comfort, expressed as the draft rate (DR), through multi-objective optimization. Numerical analysis used the standard k-ε turbulence model, validated against particle image velocimetry data, and N2O tracer transport, with duct-informed vent inlet profiles from a two-stage simulation. Four tracer species, one per occupant, released simultaneously, resolved all 12 source-receptor cross-infection pathways through a cell-wise Wells-Riley model; draft was quantified by the cabin-averaged and worst-seat DR. A Taguchi L27 array with Grey Relational Analysis (GRA) consolidated the 13 conflicting responses into a Grey Relational Grade (GRG), 90.3% of whose variance was due to inlet velocity. The predicted optimum, 9 m/s and 12°C with nominal vent angles of 60°, 0°, 60° and 60°, was confirmed by a direct simulation (GRG 0.941; predicted 0.922 ± 0.099) and by runs varying Vent 1 and Vent 2 at fixed velocity and temperature. It reduces the mean infection probability to 1.5%, 94% below the worst case, but at cabin-averaged and worst-seat DRs of 82.8% and 99.3%, far beyond ISO 7730 Category III; the same layout at 20°C has the highest grade of all runs (0.946), with a cabin-averaged DR of 67.7% at unchanged infection risk. Both are temporary high-dilution modes, not comfortable operating points.
