Indirect measurements using instrumented moving vehicles offer transformative potential for structural health monitoring of network-level bridges, providing operational efficiency and cost-effectiveness without disrupting traffic. Existing studies rely on simplified two-dimensional (2D) vehicle and bridge models, which fail to capture essential 3D dynamics of real-world systems-including aliased flexural-torsional mode shapes and spatially distributed damage across longitudinal and lateral girders-limiting their practical applicability. This study introduces a novel framework for 3D spatial mode shape reconstruction and damage localization in multi-girder bridges using an instrumented 3D two-axle vehicle. Three primary innovations are presented: (1) a 3D vehicle-bridge interaction model, integrating theoretical formulations, finite element simulations, and lab tests, to capture vertical, lateral, and torsional bridge behaviors; (2) a hybrid algorithm combining residual contact-point (CP) response, continuous wavelet transform (CWT), and modal flexibility curvature (MFC) indicator to isolate vehicle, pavement, and bridge information from vehicle accelerations; (3) the first demonstration of spatial damage localization across longitudinal and lateral girders, unachievable by conventional 2D methods. Numerical validations show that residual CP accelerations enable precise identification of 3D bridge frequencies and mode shapes under moderate road roughness. MFC indicators quantify damage severity and pinpoint spatial locations, with ongoing traffic enhancing accuracy. While residual CP responses mitigate roughness-induced noise, spectral contamination persists, limiting damage detection under significant roughness. Reconstructed 3D modes enable spatial damage identification under minimal roughness, though complete interference elimination remains unachieved. Preliminary lab tests confirm the feasibility of identifying flexural-torsional frequencies and distinguishing mode shapes, though spatial damage localization requires further refinement. This study advances the field by demonstrating the use of an instrumented 3D test vehicle to measure modal properties and structural damage in real-world 3D multi-girder bridges.
