Accurate façade-level convective heat-transfer coefficients (CHTCs) are essential for building energy and urban climate modeling. This study develops façade-specific predictive correlations for forced convection in idealized urban arrays using Reynolds-Averaged Navier–Stokes simulations. A 3 × 3 building array is analyzed while systematically varying plan area density (0.04- 0.25), height ratio (h/W = 1– 4), wind angle (θ = 0- 70 ◦ ), and reference wind speed (U 10 = 5–16 m/s). The results are validated again wind-tunnel data. In addition to quantifying façade-averaged CHTCs, the study analyzes how these parameters govern the underlying velocity field, streamline topology, and near-surface flow behavior. Results show that (i) array and façade-averaged CHTCs scale sub-linearly with wind speed; (ii) CHTCs decline with increasing plan area density due to blockage and mutual sheltering; (iii) wind angle redistributes convection among façades and (iv) taller buildings exhibit larger CHTCs, especially on windward façades and roofs, as surfaces are immersed in faster layers of ABL and experience stronger edge shear. Nonlinear multivariable regression yields closed-form correlations for windward, leeward, lateral, and top surfaces, provided for both the array average and the central (representative) building. Predictions closely track CFD, enabling façade-resolved CHTC estimation without rerunning CFD.
