How do you predict and analyze flow separation in aerodynamic design?

Flow separation prediction methods include: boundary layer integral methods (Thwaites method for laminar, Head method for turbulent), CFD with appropriate turbulence models (RANS with SA or k-omega SST), pressure gradient analysis (adverse gradients cause separation), and wind tunnel testing with flow visualization (tufts, oil flow, PIV). Key indicators are skin friction coefficient approaching zero, reversed flow near surface, and shape factor (H) exceeding critical values (~2.4 for turbulent). Separation location affects maximum lift, drag, and control effectiveness.