Precise calibration of constitutive models for cyclic liquefaction is essential but often time-consuming and requires significant expertise, limiting broader application in geotechnical practice. This paper introduces an automatic calibration tool designed to streamline the process for advanced constitutive models under both monotonic and cyclic loading. The tool supports various types of monotonic and cyclic laboratory test data, offers multiple choices of suitable comparison planes for error calculation, with a focus to also suit cyclic liquefaction problems, and employs advanced optimization techniques. The calibration follows a two-stage approach: first, optimizing parameters governing monotonic response using monotonic test data; second, refining these and additional parameters with both monotonic and cyclic data. The critical state parameters are fixed throughout, while the elasticity parameters are fixed in the second stage, all within defined bounds. Using this automatic calibration tool and the adapted calibration strategy, extensive element-level test data was used to determine the parameters of the SANISAND-MSf model for a given sand. These calibrated parameters were then used to simulate boundary value problems, including centrifuge tests of liquefiable sand slopes and sheet-pile-supported liquefiable sand deposits, all subjected to base excitations, demonstrating excellent alignment with experimental results. This validation highlights the robustness, reproducibility, and accuracy of the tool to model cyclic liquefaction while significantly reducing the expertise and time required for calibration. This represents a significant advancement toward the broader adoption of advanced constitutive soil models in geotechnical engineering practice.

Integral bridges with longer spans experience an increased cyclic interaction with their granular backfills, particularly due to seasonal thermal fluctuations. To accurately model this interaction behaviour under cyclic loading, it is crucial to employ appropriate constitutive models and meticulously calibrate and test them. For this purpose, in this paper two advanced elastoplastic (DeltaSand, Sanisand-MS) and two hypoplastic (Hypo+IGS, Hypo+ISA) constitutive models with focus on small strain and cyclic behaviour are investigated. The soil models are calibrated based on a comprehensive laboratory programme of a representative highly compacted gravel backfill material for bridges. The calibration procedure is shown in detail and the model capabilities and limitations are discussed on the element test level. Additional triaxial tests with repeated un- and reloading reveal significant over- and undershooting effects for the majority of the investigated material models. Finally, cyclic finite element analyses on the soil-structure interaction of an integral bridge are conducted to compare the performance of the soil models. Qualitatively similar cyclic evolution of earth pressures are detected for the soil models at various bridge lengths and test settings. However, a substantially different cyclic settlement behaviour is observed. Additionally, the investigation highlights severe overshooting effects associated with the tested hypoplastic soil models. This phenomenon is studied in detail using a single integration point analysis. Supplementary studies reveal that the foot point deformation of the abutment significantly influences the lateral passive stress mobilisation and the amount of its increase with growing seasonal cycles.