Abstract

Estuarine tidal flats represent a critical environment for maintaining healthy ecosystems. The geotechnical characterization of such environments is often challenging due to their soft nature and remote locations. Thus, the goal of this study is to explore the use of remotely sensed, high resolution (pixel resolution of ~0.5 m) synthetic aperture radar (SAR) data for characterizing these sediments. Towards this goal, the variability of sediment properties (median gain size, moisture content, fines content, and classification following the Unified Soil Classification System (USCS) were documented for a mixed tidal flat located on the Great Bay Esturay in New Hampshire, USA. A sediment classification scheme based on SAR, originally proposed by Gade et al. (2008), which utilizes the correlation length to characterize sediments based on the percentage of particles <0.063 mm, was tested using a high-resolution X-band SAR image (pixels of ~0.5 m) collected simultaneously to in situ measurements and a medium-resolution C-band image (pixels of ~10 m) collected 3 days later. The results of this study demonstrate that the framework proposed by Gade et al. (2008) was generally able to predict the correct soil classification for coarser sediments but failed to predict the correct soil classification for finer sediments when applying the original image resolutions. When the X-band image was downsampled to match the C-band image resolution, the framework was still unable to predict the expected soil classification. This suggests that, in order to utilize high-resolution X-band data to predict soil type, alternative sediment type classification schemes are required.

Abstract

The use of soil data is essential in geotechnical design, but in a preliminary project phase such data are usually limited to that inferred from field tests, like CPT, SPT or DMT. In previous publications by the authors and co-workers, it was shown how such data can be automatically processed into soil profiles and parameter sets for geotechnical finite element analysis. Another publication demonstrated the automated processing and creation of geological models as an intermediate step to more advanced 3D geotechnical modelling in a BIM / Digital Twin environment, which facilitates the link with other disciplines and stakeholders in a project. The major challenge of connecting layers across multiple 1D boreholes to form 3D soil layers is overcome by using a Machine Learning clustering algorithm. As a next step, the previously introduced Automated Parameter Determination (APD) method (connecting correlations using Graph theory) is applied based on averaged CPT parameters from all contributing layer sections. The result is an automated system that creates a complete 2D or 3D finite element model, including constitutive model parameters, for geotechnical analysis purposes. An automated system may be very efficient when exploring different design alternatives in an early stage of a project. However, it is important to emphasize the role and responsibilities of the geotechnical engineer in the design process, which requires the system to be transparent, verifiable, and adaptable. This paper describes the state-of-the-art of this ongoing research project.

Abstract

Soil boundary delineation is an important task in geotechnical site characterization. It can be achieved by either extracting borehole samples, conducting laboratory tests, and classifying them according to a soil classification system such as the Unified Soil Classification System (USCS) or utilizing multiple cone penetration test (CPT) soundings, and identifying soil boundaries at the soundings from the Ic (soil behavior type index) profiles. However, most soil-layer delineation methods can only take a single type of test result as the input. For instance, the well-known Markov random field (MRF) method can only take soil-type data such as sand, silt, or clay at boreholes as the input. Recognizing that soil classifications and soil properties are correlated, this paper proposes a novel coupled MRF-Bayesian framework to infer the spatial variation of USCS classifications (e.g., sand, silt, and clay) as well as soil properties by integrating both CPT and borehole data. This integrated approach leverages both CPT and borehole data to address some main challenges e.g., uncertainties and multivariate soil data input in underground stratification problems by simultaneous sampling of soil properties and soil types. The new unified framework can accommodate multivariate data, hence the new framework is compatible with the geotechnical engineering practice. The uncertainties for the spatial variation of USCS classification at sounding locations are quantified through a “layer-specific” Bayesian updating i.e., updating posterior cross-correlation behaviors for different layers (such as sand, silt, and clay), independently. In this Bayesian updating, soil-type data can provide some information about the soil properties according to the unified soil classification system. Further, the soil boundaries can be identified across the entire domain by the realization of conditional random fields of soil properties once the spatial variation of USCS classification is inferred at sounding locations, followed by a 3-dimensional Markov random field process.

Abstract

Soil boundary delineation is an important task in geotechnical site characterization. It can be achieved by either extracting borehole samples, conducting laboratory tests, and classifying them according to a soil classification system such as the Unified Soil Classification System (USCS) or utilizing multiple cone penetration test (CPT) soundings, and identifying soil boundaries at the soundings from the Ic (soil behavior type index) profiles. However, most soil-layer delineation methods can only take a single type of test result as the input. For instance, the well-known Markov random field (MRF) method can only take soil-type data such as sand, silt, or clay at boreholes as the input. Recognizing that soil classifications and soil properties are correlated, this paper proposes a novel coupled MRF-Bayesian framework to infer the spatial variation of USCS classifications (e.g., sand, silt, and clay) as well as soil properties by integrating both CPT and borehole data. This integrated approach leverages both CPT and borehole data to address some main challenges e.g., uncertainties and multivariate soil data input in underground stratification problems by simultaneous sampling of soil properties and soil types. The new unified framework can accommodate multivariate data, hence the new framework is compatible with the geotechnical engineering practice. The uncertainties for the spatial variation of USCS classification at sounding locations are quantified through a “layer-specific” Bayesian updating i.e., updating posterior cross-correlation behaviors for different layers (such as sand, silt, and clay), independently. In this Bayesian updating, soil-type data can provide some information about the soil properties according to the unified soil classification system. Further, the soil boundaries can be identified across the entire domain by the realization of conditional random fields of soil properties once the spatial variation of USCS classification is inferred at sounding locations, followed by a 3-dimensional Markov random field process.

Abstract

Cemented geomaterials exist in many parts of the world. Structure and bonding largely influence their strength, stiffness, permeability and other hydromechanical properties. Despite the CPT being the most widely used geotechnical engineering soil characterization tool, most existing correlations between penetration resistance and soil parameters apply only to uncemented granular deposits. Application of existing correlations to cemented geomaterials such as soft rocks can produce misleading interpretation making CPT application more challenging. In particular, CPT databases correlating tip resistance with yield stress of the intact material show a wide scatter prompting the need for a better understanding of the mechanics of cone penetration in soft bonded materials. In this work, 1g small-scale model CPTs are performed in a soft rock, whilst in-test X-ray techniques help to reveal mechanisms behind the penetration process. Thereafter, experimental results are compared to field scale results and those modelled using the Particle Finite Element method which is geared toward large deformation analyses. The combined interpretation of the experimental and numerical data is then used to discuss some of the unique attributes of CPT behaviour in soft rock.

Abstract

Wave energy converters (WECs), a form of marine hydrokinetic (MHK) energy device, transform the mechanical energy of water waves into electricity. They are typically held on station using anchoring systems embedded into the seafloor sediments (as opposed to, e.g., concrete gravity anchors). The design of WEC anchors is particularly challenging for two primary reasons: (1) WECs are often deployed in previously undeveloped areas of the ocean, so the engineering properties of the seabed are largely unknown; and (2) the economic margins on wave energy are quite thin, so heavily overdesigned anchor systems in response to data sparsity are not feasible. This paper describes the planning, execution, and outcomes from a dedicated in-situ testing campaign informed only by limited geophysical data a priori. A series of 22 cone penetration tests (CPTs) were performed at a 7-km2 site approximately 11 km off the United States’ west coast. Water depth was up to approximately 75 m and the target depth for the cone soundings was 10.5 m below the seafloor. Measurements indicated that much of the site subsurface consisted of dense sand and gravel, though tests identified an overlying softer layer in some parts of the site. A summary of lessons learned and recommendations for future explorations at similarly unexplored sites are provided.

Abstract

One of the ways to enhance the efficiency of the cone penetration testing process is to mount modules behind the cone. In this way the test will not only generate the standard Cone Penetration Testing (CPT) data (i.e., cone tip resistance, sleeve friction, and dynamic pore water pressure), but also the data obtained by the module pushed into the soil together with the cone. While for certain modules it is common practice to analyze the acquired data extensively (e.g., the seismic module) for other modules this is not necessarily the case. A good example of the latter is the video module, which has been available for several decades. When this module is deployed with visible light, the analysis is typically limited to viewing the recording and adding observation notes. During the recent TRIM4 research project the video module was deployed and subsequently attempts were made to identify the soil type through an automatic analysis of the video images and to characterize and to determine the grain size distribution using the video images. This approach is highly correlated with the soil behavior type index, commonly used in the analysis of CPT data, and at the same time mitigates the effect of the CPT data reflecting changes in soil strength behavior before a layer is actually penetrated by the cone. In this paper the authors will describe the use of the video cone in very general terms, but focus on this analysis methodology in detail