Advanced models for soft organic layers encountered in the shallow subsoils in the Netherlands have been developed recently at TU Delft. The models are based on high-quality laboratory data on peats and soft organic clays. The constitutive effort mostly focussed on some partially unexplored features, such as the role of fibres, extension stress conditions, and the dependence of hardening on deviatoric plastic strains, besides anisotropy. Although the models have proven to be able to reproduce and predict the behaviour over a variety of triaxial probe tests, validation at the field scale is lagging behind. On the one hand, field soil response encompasses diverse stress paths and histories not replicable in the laboratory. On the other hand, the role of the advanced features introduced in the models on the engineering structure response needs to be quantified. We back-analyse a well-documented full-scale test performed in the Netherlands, the Leendert de Boerspolder stress test, where the role of different soft soil layers both on the pre-failure and failure response has been investigated. Comparison between numerical simulations and available monitoring data is used to demonstrate the contribution of advanced models to the understanding of the engineering response of soft soils.@en

In the Netherlands, natural gas extraction from the Groningen field has induced seismic activities, notably recording the highest magnitude of ML 3.6 at Huizinge. Despite the lower magnitudes of these induced earthquakes compared to their natural counterparts, their impact is significantly magnified due to the shallow depths at which they occur and the high site amplification over areas with soft soils. In response, the Dutch Research Council (NWO) launched the DeepNL research programme to conduct thorough scientific research into these phenomena. The SOFTTOP project, part of this initiative, aims to fill the knowledge gaps regarding the effects of soft soil deposits on earthquake dynamics and subsequent soil responses. This thesis, nested within the SOFTTOP project, focuses on the cyclic and dynamic behaviours of soft organic clays, which are prevalent in the Netherlands' deltaic subsurface. It starts with a detailed investigation into the monotonic behaviours of these clays, laying the groundwork for further exploration into their cyclic and rate-dependent responses.
Through an extensive experimental programme, a characteristic organic diatomaceous clay from the Netherlands was tested under various stress paths, uncovering limitations in existing models. To address these, an advanced elastoplastic model, JMC-clay, was developed. Starting from a previous formulation for peats, modifications were made to the yield surface shape and the rotational hardening rule, enhancing the models' ability to predict pre-failure behaviours accurately. This foundational work paves the way for studying cyclic behaviour under ‘slow’ testing conditions, deliberately chosen to minimise the effects of non-uniform pore pressure distribution. The study's exploration into the slow cyclic response of Dutch organic clay has unearthed critical insights into the effects of loading frequency, soil composition, initial stress state, and cyclic stress amplitude on the soil's mechanical properties. Among the findings, the soil's cyclic response significantly depends on the cyclic strain amplitude.
A pivotal aspect of this thesis is extending the JMC-clay model to include cyclic behaviour, incorporating bounding surface plasticity for a more accurate and predictive modelling framework of soil behaviour under cyclic loading conditions. This development, however, brought to light challenges in model validation. Simulations indicated that a larger bounding surface, indicative of a higher apparent over-consolidation ratio (OCR), aligns more closely with experimental observations than anticipated. This suggests that contrary to the expected purely elastoplastic response, creep behaviour plays a significant role during consolidation, necessitating adjustments to the model to capture these observations accurately. The JMC-clay model is extended to include an elastoplastic-viscoplastic bounding surface formulation to capture time-dependent soil response. The strain-rate saturation feature of the coupled elastoplastic-viscoplastic framework requires further investigation with experimental data at a higher loading rate.
However, the existing equipment falls short when conducting ‘fast’ testing, designing and constructing a cyclic-dynamic multi-directional shear apparatus for organic soft soils (CYC-DoSS) in-house aimed to fully cover the earthquake frequency spectrum. The development and application of the device, a state-of-the-art earthquake simulator, marks a significant leap in element testing. Designed to overcome the limitations of traditional testing apparatuses, the CYC-DoSS features digitally controlled servo-hydraulic actuators and in-house developed local response sensors, offering a comprehensive suite of measurement capabilities. The inclusion of advanced measurement techniques, such as P-wave and S-wave bender element measurements, accelerometers, and fibre optic pore pressure transducers, alongside custom-developed sensors for capturing detailed deformation patterns, pore pressure responses, and acceleration data, significantly enhances the ability to measure cyclic responses. This is crucial for accurate seismic risk assessment and mitigation strategies and enables a deeper insight into the dynamic behaviours of soft soils under seismic load.
By weaving together experimental research, advanced constitutive modelling, and the deployment of an innovative testing apparatus, this thesis presents a comprehensive approach to unravelling the behaviour of Dutch soft organic clays under seismic conditions. The contributions of this research extend beyond theoretical advancements, offering practical insights and methodologies to enhance the resilience of infrastructure in seismic-prone areas.@en

Element testing of soft soils is challenging due to the large strains attained in the pre-failure range. Besides the heterogeneity of natural samples, the set-up configuration is the main driving factor for non-homogenous response. Stress, strain and pore pressure non-uniformities induced by the loading system affect the observed behaviour and complicate proper interpretation of the results. Among the difficulties encountered in the interpretation of laboratory data, the unexpected decrease of the stress ratio frequently observed on Dutch organic soft clays on the wet side of critical state is investigated by numerically back-analysing the triaxial test set-up. A 3D finite element simulation using an advanced constitutive model for soft clays developed at TU Delft was performed to clarify the nature of the response. The results indicate that a decrease in the deviatoric stress up to critical state may be interpreted as a true feature of the soil response. However, the response at large strains is very much influenced by the triaxial shear apparatus, in particular, by the rotation of the top cap which triggers geometrical instability. Practical recommendations are given to limit the effects of the set-up configuration on the determination of the undrained shear strength to be used for field applications.@en

Increasing frequency and intensity of extreme weather events in the Netherlands is raising attention on the unsaturated response of geo-infrastructures, promoting research projects to provide an overview of the impact of unsaturated conditions on the response of shallow soil layers and embankments, and to better address maintenance and mitigation measures. As part of this effort, we discuss the results of standard laboratory tests performed on initially unsaturated samples retrieved from the field and tested in natural conditions, as well as after controlled drying and wetting. The variation of the "undrained"(i.e. at constant water content) shear strength with the degree of saturation obtained from the laboratory tests aligns well with CPT measurements performed in the field. An elastic-plastic constitutive model with mixed isotropic-rotational hardening developed for saturated soft soils was extended to unsaturated conditions by following a robust approach previously developed for compacted clayey soils. Coupling between the mechanical and the hydraulic behaviour is provided by the water retention curve. The model nicely captures the response observed in the laboratory, until extreme dry conditions, which possibly alter the structure of the soil, the peak stress, and the brittleness after failure. The model is capable of reproducing the effects of the previous hydraulic history on the stress-strain behaviour observed from the laboratory tests over a wide range of degree of saturation.

@en

Recent research effort carried out at Delft University of Technology to improve the experimental knowledge and develop a comprehensive modelling approach for fibrous organic soils is summarised. Experimental results and numerical analyses are combined to discuss some contradictory results which have delayed advanced characterisation of peats. Part of the apparent inconsistencies commonly found in the literature is due to the influence of the testing apparatus, including rough platens and membrane restraint, which inhibit homogenous deformation modes and alter the response of the samples compared to the true material behaviour. The consequences of non-homogenous deformation are particularly relevant on peats due to the unique combination of their exceptionally low stiffness and high strength. An elastic–plastic constitutive framework was developed starting from repeatable reconstituted samples of peats, taking care of reducing end restraint to a large extent in the experimental setup. The results suggested that an elastic–plastic model for peats should include a non-associated flow rule and a mixed volumetric–deviatoric hardening law. The role played by different fibres at the laboratory scale is discussed, and the additional reinforcement offered by bigger fibres on the observed behaviour of natural peats is addressed.

@en

A simple numerical model was set up to investigate the hydraulic behaviour of a regional dyke to improve understanding of the response under variable atmospheric conditions. The unsaturated hydraulic properties of the dyke body and the relevant foundation layers were calibrated either on the results of laboratory tests or on a national database, namely the Staringreeks, compiled for typical Dutch soils. The boundary conditions were imposed according to the weather history at the top, and to the pore pressures measured in the field at the bottom of the representative soil column. The results indicate that a simple 1D model is able to accurately reproduce the suction time history in the dyke core, provided the hydraulic conductivity and soil water retention properties are properly calibrated. The optimised hydraulic conductivities are typically two orders of magnitude higher than the saturated hydraulic conductivity from the laboratory tests, but comparable to the ones suggested in the database developed on field data. The work highlights that cautious evaluation of laboratory data is needed for field applications, and that direct information from the field should be used to validate numerical models in the presence of organic soils.

@en