Comparative Performance of Micropile and Conventional Pile Retrofits: Strain-Dependent Head Stiffness from 1g Tests and Numerical Analysis

Seismic retrofitting of bridge foundations in urban environments must satisfy stringent performance requirements while contending with spatial constraints that preclude adding large-diameter piles. Micropiles offer a compact alternative, yet their soil-structure interaction (SSI) behaviour relative to conventional pile-addition remains insufficiently quantified across strain levels. This study combines nonlinear numerical modelling with scaled 1g model testing to compare conventional pile-addition retrofits against micropile retrofits. A baseline un-retrofitted 2×2 floating pile group, a conventional pile-addition scheme, and micropile retrofit configurations were evaluated to obtain head-level lateral stiffness under controlled cyclic displacements, with explicit treatment of soil nonlinearity and soil-pile interface effects. Physical 1g model tests measured head-level impedance functions (IFs) for a 2×2 pile group retrofitted with micropiles at 0°, 5°, and 10° inclinations; loading amplitudes were varied to consider progressive nonlinearity in the surrounding soil. At low strains, micropiles (particularly when inclined) produce higher head-level stiffness than the un-retrofitted baseline. At higher strains, nonlinearity in the surrounding soil reduces this stiffness advantage, with micropile performance approaching or falling below that of the un-retrofitted baseline. The conventional retrofit, on the other hand, shows better performance than micropiles at such high strain levels. The convergent trends between physical model tests and numerical predictions provide validated direction in selecting retrofit configurations based on expected loading intensities in space-constrained sites, clarifying the strain-level regimes where micropiles offer optimal seismic performance.