R.B.J. Brinkgreve
info
Please Note
<p>This page displays the records of the person named above and is not linked to a unique person identifier. This record may need to be merged to a profile.</p>
68 records found
1
This paper demonstrates a comprehensive workflow for probabilistic and sensitivity analysis of settlement of a hypothetical embankment case using realistic CPT data. The stratification reveals layers with different behaviour types, including clays, silt mixtures, sandy mixtures and sands. The dominant layer consists of soft clay which is susceptible to excessive deformation due to the embankment construction. Given the significant variations in the CPT readings, deformation analysis with deterministic parameters is insufficient. Instead, a series of finite element analyses were conducted with different material properties. The soft clay was modeled using the Soft Soil constitutive model, while other layers were modeled using the Hardening Soil Small Strain model in PLAXIS 2D. Model parameters were randomly selected from the parameter sets derived from individual CPT readings of the respective layer. In this way, each reading was considered as an independent measurement instance, with consistent derived parameters. Uniform sampling over depth maintained the actual distribution of the parameters within each layer. Additionally, the unit weight of the embankment was randomly chosen from its probability density function. Latin hypercube sampling was employed to ensure even sampling across the parameter space with a smaller number of samples. The resulting embankment settlements were processed in terms of the probability of exceeding a predefined maximum allowed value. To identify the significant contributing factors to settlement and deformation variation, Sobol’ global sensitivity analysis was performed. Uncertain parameters included representative features for each layer and the embankment's unit weight. The Sobol’ procedure requires considerably more simulations than required by the probabilistic analysis, which is prohibitive because of the computational demands. Therefore, several surrogate models were trained using machine and deep learning algorithms. To achieve acceptable performance, hyperparameters were tuned automatically over appropriate ranges. The best-scoring model was then employed to estimate the embankment displacements in the sensitivity analysis.
...
This paper demonstrates a comprehensive workflow for probabilistic and sensitivity analysis of settlement of a hypothetical embankment case using realistic CPT data. The stratification reveals layers with different behaviour types, including clays, silt mixtures, sandy mixtures and sands. The dominant layer consists of soft clay which is susceptible to excessive deformation due to the embankment construction. Given the significant variations in the CPT readings, deformation analysis with deterministic parameters is insufficient. Instead, a series of finite element analyses were conducted with different material properties. The soft clay was modeled using the Soft Soil constitutive model, while other layers were modeled using the Hardening Soil Small Strain model in PLAXIS 2D. Model parameters were randomly selected from the parameter sets derived from individual CPT readings of the respective layer. In this way, each reading was considered as an independent measurement instance, with consistent derived parameters. Uniform sampling over depth maintained the actual distribution of the parameters within each layer. Additionally, the unit weight of the embankment was randomly chosen from its probability density function. Latin hypercube sampling was employed to ensure even sampling across the parameter space with a smaller number of samples. The resulting embankment settlements were processed in terms of the probability of exceeding a predefined maximum allowed value. To identify the significant contributing factors to settlement and deformation variation, Sobol’ global sensitivity analysis was performed. Uncertain parameters included representative features for each layer and the embankment's unit weight. The Sobol’ procedure requires considerably more simulations than required by the probabilistic analysis, which is prohibitive because of the computational demands. Therefore, several surrogate models were trained using machine and deep learning algorithms. To achieve acceptable performance, hyperparameters were tuned automatically over appropriate ranges. The best-scoring model was then employed to estimate the embankment displacements in the sensitivity analysis.
Conference paper
(2026)
-
Ricardo J. N. Azeiteiro, Ilaria Del Brocco, Evira Eman, Ronald B. J. Brinkgreve, Tuan A. Bui, Sandro Brasile
The Hardening Soil (HS) model has been widely used for the simulation of geotechnical problems. Its popularity has been sustained by its ability to simulate some of the key features of soil response, such as stress dependency, dilatancy and highly non-linear pre-failure response, associated with its relatively simple and clear calibration process based on standard laboratory and field test results and well-established empirical formulas. To improve its ability to deal with unloading, the model was later extended with a small-strain stiffness formulation, leading to the HS-small model. It has become a natural choice for simulations involving retaining walls, shafts and tunnels. Moreover, given the hysteretic nature of the small-strain stiffness formulation, the HS-small model has also emerged as a suitable choice for the simulation of wave propagation and vibration problems within the small to medium strain range. However, as many other constitutive models based on non-linear hysteretic elasticity (or plasticity), it has been recently found that HS-small can suffer from overshooting issues, which are characterised by an overprediction of stiffness after the occurrence of minor unloading-reloading cycles, due to the reset of the strain history tensor. Besides being often found in dynamic problems (e.g. vibrations induced by pile driving, traffic loading or machine operation), these issues may generally affect all boundary-value problem simulations, where numerical oscillations often produce spurious unloading-reloading perturbances. It is therefore imperative to address them. In this paper, the fundamental ingredients of two novel small-strain stiffness formulations addressing overshooting issues in practical applications are presented. The results of laboratory test simulations, including small unloading-reloading cycles interpreted as perturbances, show that both proposed formulations are able to prevent overshooting issues in HS-small.
...
The Hardening Soil (HS) model has been widely used for the simulation of geotechnical problems. Its popularity has been sustained by its ability to simulate some of the key features of soil response, such as stress dependency, dilatancy and highly non-linear pre-failure response, associated with its relatively simple and clear calibration process based on standard laboratory and field test results and well-established empirical formulas. To improve its ability to deal with unloading, the model was later extended with a small-strain stiffness formulation, leading to the HS-small model. It has become a natural choice for simulations involving retaining walls, shafts and tunnels. Moreover, given the hysteretic nature of the small-strain stiffness formulation, the HS-small model has also emerged as a suitable choice for the simulation of wave propagation and vibration problems within the small to medium strain range. However, as many other constitutive models based on non-linear hysteretic elasticity (or plasticity), it has been recently found that HS-small can suffer from overshooting issues, which are characterised by an overprediction of stiffness after the occurrence of minor unloading-reloading cycles, due to the reset of the strain history tensor. Besides being often found in dynamic problems (e.g. vibrations induced by pile driving, traffic loading or machine operation), these issues may generally affect all boundary-value problem simulations, where numerical oscillations often produce spurious unloading-reloading perturbances. It is therefore imperative to address them. In this paper, the fundamental ingredients of two novel small-strain stiffness formulations addressing overshooting issues in practical applications are presented. The results of laboratory test simulations, including small unloading-reloading cycles interpreted as perturbances, show that both proposed formulations are able to prevent overshooting issues in HS-small.
This study presents a comprehensive numerical investigation into the use of displacement piles as a reinforcement measure for river dikes founded on soft soil, with a particular focus on geotechnical performance, macro stability, and impacts on nearby buildings. A finite element model is developed using parameters derived from a representative Dutch dike case (Bergambacht), incorporating the Hardening Soil, Soft Soil Creep and NGI-ADP-SHANSEP models to capture soil behaviour. Pile installation is simulated through the application of lateral volumetric strain, with varying pile diameters, spacings, and locations within the dike profile. The equivalent diameters used in the analysis range from 10 to 40 cm, corresponding to pile walls with diameters between 25.5 and 100 cm when the spacing equals the diameter. The pile wall location varies from the dike toe up to 21 m away, which is at the outer crest, with a varied length reaching -12 m NAP. A two-storey building on deep pile foundations is included to assess the effect of installation-induced displacements, with its location ranging from 5 to 20 m from the dike toe. Results show that positioning the pile wall within the inner slope offers the best balance between increased factor of safety, reduced required pile length, and acceptable levels of deformation. However, the installation process can generate significant horizontal displacements, particularly near the dike toe, which may compromise adjacent structures. The study finds that displacement piles are unsuitable within 10–15 m of existing buildings unless smaller pile diameters or alternative installation methods are used. Soil stiffness and installation-induced stresses also play a key role, highlighting the importance of site-specific assessments and careful design calibration using field data.
...
This study presents a comprehensive numerical investigation into the use of displacement piles as a reinforcement measure for river dikes founded on soft soil, with a particular focus on geotechnical performance, macro stability, and impacts on nearby buildings. A finite element model is developed using parameters derived from a representative Dutch dike case (Bergambacht), incorporating the Hardening Soil, Soft Soil Creep and NGI-ADP-SHANSEP models to capture soil behaviour. Pile installation is simulated through the application of lateral volumetric strain, with varying pile diameters, spacings, and locations within the dike profile. The equivalent diameters used in the analysis range from 10 to 40 cm, corresponding to pile walls with diameters between 25.5 and 100 cm when the spacing equals the diameter. The pile wall location varies from the dike toe up to 21 m away, which is at the outer crest, with a varied length reaching -12 m NAP. A two-storey building on deep pile foundations is included to assess the effect of installation-induced displacements, with its location ranging from 5 to 20 m from the dike toe. Results show that positioning the pile wall within the inner slope offers the best balance between increased factor of safety, reduced required pile length, and acceptable levels of deformation. However, the installation process can generate significant horizontal displacements, particularly near the dike toe, which may compromise adjacent structures. The study finds that displacement piles are unsuitable within 10–15 m of existing buildings unless smaller pile diameters or alternative installation methods are used. Soil stiffness and installation-induced stresses also play a key role, highlighting the importance of site-specific assessments and careful design calibration using field data.
Automated parameter determination
From in-situ measurements to constitutive models
Site characterization relies on both in-situ and laboratory testing. In the early stages of a project, in-situ tests are often performed before launching a full laboratory testing program. At this stage—when soil data is limited—in-situ tests can provide valuable insights for preliminary characterization. To enhance the interpretation of these tests, an automated parameter determination framework has been developed, employing a graph-based approach to derive soil and constitutive model parameters from in-situ measurements. Several studies have been conducted to validate the framework’s output in terms of both soil properties and model parameters. The framework is designed to be transparent and adaptable, allowing users to trace the computed values for different parameters and incorporate their experience, knowledge and expertise. In this study, the tool was applied to a well-documented test site in Australia. Additionally, the integration of machine learning models for predicting soil parameters is explored as part of ongoing efforts to incorporate data-driven techniques into the framework.
...
Site characterization relies on both in-situ and laboratory testing. In the early stages of a project, in-situ tests are often performed before launching a full laboratory testing program. At this stage—when soil data is limited—in-situ tests can provide valuable insights for preliminary characterization. To enhance the interpretation of these tests, an automated parameter determination framework has been developed, employing a graph-based approach to derive soil and constitutive model parameters from in-situ measurements. Several studies have been conducted to validate the framework’s output in terms of both soil properties and model parameters. The framework is designed to be transparent and adaptable, allowing users to trace the computed values for different parameters and incorporate their experience, knowledge and expertise. In this study, the tool was applied to a well-documented test site in Australia. Additionally, the integration of machine learning models for predicting soil parameters is explored as part of ongoing efforts to incorporate data-driven techniques into the framework.
Dada la creciente complejidad del entorno de construcción en zonas urbanas densamente pobladas y condiciones de terreno problemáticas, el diseño de túneles y lumbreras supone todo un reto. Comprender el comportamiento del suelo y de la roca y la interacción con las estructuras es esencial a la hora de realizar análisis numéricos como parte del proceso de diseño. Pero hay más: el diseño de proyectos complejos requiere en la actualidad de un enfoque multidisciplinario basado en una única fuente de verdad para todos los datos implicados. Este artículo destaca algunos de los avances recientes en la modelización numérica de túneles y pozos en suelo y roca. Demuestra cómo el análisis numérico está integrado en un entorno basado en datos y cómo la modelización y la interpretación de datos pueden beneficiarse de la automatización y la visualización en 3D del subsuelo. El resultado es un flujo de trabajo eficaz y menos propenso a errores, que da lugar a diseños más económicos con menor riesgo de fallas. Por último, se ofrece una perspectiva sobre la manera en la que los datos dominarán el futuro proceso de diseño, mediante el uso de inteligencia artificial (AI) y el aprendizaje automático (ML).
...
Dada la creciente complejidad del entorno de construcción en zonas urbanas densamente pobladas y condiciones de terreno problemáticas, el diseño de túneles y lumbreras supone todo un reto. Comprender el comportamiento del suelo y de la roca y la interacción con las estructuras es esencial a la hora de realizar análisis numéricos como parte del proceso de diseño. Pero hay más: el diseño de proyectos complejos requiere en la actualidad de un enfoque multidisciplinario basado en una única fuente de verdad para todos los datos implicados. Este artículo destaca algunos de los avances recientes en la modelización numérica de túneles y pozos en suelo y roca. Demuestra cómo el análisis numérico está integrado en un entorno basado en datos y cómo la modelización y la interpretación de datos pueden beneficiarse de la automatización y la visualización en 3D del subsuelo. El resultado es un flujo de trabajo eficaz y menos propenso a errores, que da lugar a diseños más económicos con menor riesgo de fallas. Por último, se ofrece una perspectiva sobre la manera en la que los datos dominarán el futuro proceso de diseño, mediante el uso de inteligencia artificial (AI) y el aprendizaje automático (ML).
Discussion Session D2 (Embankments)
Contribution of the Discussion Leader
Embankments are special structures that play a crucial role in various civil engineering projects, namely in the construction of roads, railways, dams, dikes, airports, and more. These types of earthworks provide vital support, stability, and protection in diverse infrastructure developments, making them a key subject in the field of geotechnical engineering. The design and construction of embankments require careful consideration of soil properties, foundation conditions, loading requirements, and environmental factors to ensure their long-term performance and structural integrity. Embankments are not only fundamental for creating elevated platforms for transportation networks but also serve as barriers for flood protection, reservoir containment, and land reclamation. The success of embankment projects relies on effective engineering solutions that address challenges such as settlement control, slope stability, and material selection. Innovations in geotechnical practices, including the use of geosynthetics, reinforcement techniques, and sustainable construction methods, have significantly advanced the design and construction of embankments, enhancing their durability and performance. This session about embankments focuses on the latest developments, research findings, and practical applications related to embankments in geotechnical engineering. By exploring topics such as cost-effective solutions, sustainable construction practices, innovative reinforcement systems, and performance evaluation methods, this session aims to provide valuable insights and knowledge exchange opportunities for professionals, researchers, and practitioners in the field.
...
Embankments are special structures that play a crucial role in various civil engineering projects, namely in the construction of roads, railways, dams, dikes, airports, and more. These types of earthworks provide vital support, stability, and protection in diverse infrastructure developments, making them a key subject in the field of geotechnical engineering. The design and construction of embankments require careful consideration of soil properties, foundation conditions, loading requirements, and environmental factors to ensure their long-term performance and structural integrity. Embankments are not only fundamental for creating elevated platforms for transportation networks but also serve as barriers for flood protection, reservoir containment, and land reclamation. The success of embankment projects relies on effective engineering solutions that address challenges such as settlement control, slope stability, and material selection. Innovations in geotechnical practices, including the use of geosynthetics, reinforcement techniques, and sustainable construction methods, have significantly advanced the design and construction of embankments, enhancing their durability and performance. This session about embankments focuses on the latest developments, research findings, and practical applications related to embankments in geotechnical engineering. By exploring topics such as cost-effective solutions, sustainable construction practices, innovative reinforcement systems, and performance evaluation methods, this session aims to provide valuable insights and knowledge exchange opportunities for professionals, researchers, and practitioners in the field.
Numerische Ermittlung von Baugrundschwingungen bei dynamisch belasteten Fundamenten
Empfehlungen zur Modellierung
Journal article
(2024)
-
Winfried Schepers, Ronald B. J. Brinkgreve, Kira Holtzendorff, Dirk Wegener, Silke Appel, Georgia Efthymiou, Wolfgang Krajewski, Jan Machaček, Thomas Meier, More authors...
Geotechnical engineers are increasingly concerned with wave propagation problems. Manufacturers of geotechnical analysis software added features for soil dynamic analyses to their products initially devised for static geotechnical analyses. Though, users often lack the experience for conducting such advanced numerical analyses. Working groups 1.4 “Soil dynamics” and 1.6 “Numerical analyses in geotechnical engineering” of DGGT German Society for Geotechnical Engineering established a joint subgroup “Numerical analyses in soil dynamics” to address this shortcoming. The present paper presents the work of the subgroup so far and provides some guidance on conducting numerical analyses in soil dynamics.
...
Geotechnical engineers are increasingly concerned with wave propagation problems. Manufacturers of geotechnical analysis software added features for soil dynamic analyses to their products initially devised for static geotechnical analyses. Though, users often lack the experience for conducting such advanced numerical analyses. Working groups 1.4 “Soil dynamics” and 1.6 “Numerical analyses in geotechnical engineering” of DGGT German Society for Geotechnical Engineering established a joint subgroup “Numerical analyses in soil dynamics” to address this shortcoming. The present paper presents the work of the subgroup so far and provides some guidance on conducting numerical analyses in soil dynamics.
APD
An automated parameter determination system based on in-situ tests
In-situ testing has numerous applications in geotechnical engineering. The interpretation of in-situ test results includes soil stratification and determination of soil parameters. This paper presents an automated parameter determination framework that aims to determine constitutive model parameters based on in-situ tests. The ongoing research project relies on a graph-based approach for determining the parameters. The framework has two main attributes: transparency and adaptability. Transparency is achieved by illustrating how a certain parameter was computed. Adaptability is ensured by allowing users to incorporate their expertise into the framework. The system currently determines parameters based on three main workflows that utilize the results of cone penetration tests, dilatometer tests, and shear wave velocity measurements. This study employs the three main workflows to determine soil parameters for one of the Norwegian GeoTest Sites. Additionally, the connection between the parameter determination system and finite element analysis is discussed, where the parameters for the Modified Cam Clay model are evaluated. The framework is valuable in the early stages of projects, providing detailed soil information when soil data is limited. Ongoing research aims to assess the accuracy of the derived soil and constitutive model parameters and to expand the system’s capabilities by including additional in-situ tests.
...
In-situ testing has numerous applications in geotechnical engineering. The interpretation of in-situ test results includes soil stratification and determination of soil parameters. This paper presents an automated parameter determination framework that aims to determine constitutive model parameters based on in-situ tests. The ongoing research project relies on a graph-based approach for determining the parameters. The framework has two main attributes: transparency and adaptability. Transparency is achieved by illustrating how a certain parameter was computed. Adaptability is ensured by allowing users to incorporate their expertise into the framework. The system currently determines parameters based on three main workflows that utilize the results of cone penetration tests, dilatometer tests, and shear wave velocity measurements. This study employs the three main workflows to determine soil parameters for one of the Norwegian GeoTest Sites. Additionally, the connection between the parameter determination system and finite element analysis is discussed, where the parameters for the Modified Cam Clay model are evaluated. The framework is valuable in the early stages of projects, providing detailed soil information when soil data is limited. Ongoing research aims to assess the accuracy of the derived soil and constitutive model parameters and to expand the system’s capabilities by including additional in-situ tests.
Understanding and properly simulating discontinuity mechanical behaviour is crucial in all rock engineering projects. Several constitutive relationships have been proposed and implemented in numerical codes. This paper discusses the results of a numerical study that exam-ines the influence of adopting different rock discontinuity constitutive models for simulating the behavior of a fractured rock mass. Two constitutive approaches are employed: an enhanced Cou-lomb-based criterion with strain softening and a modified version of the Barton-Bandis model to overcome potential implementation issues. These models have been implemented in PLAXIS and their performance is inspected through numerical analyses of an underground cavity for a specific discontinuity network geometry. The results provide insights into the implications and suitability of adopting different discontinuity constitutive models for assessing the stability of engineering works in fractured rock masses.
...
Understanding and properly simulating discontinuity mechanical behaviour is crucial in all rock engineering projects. Several constitutive relationships have been proposed and implemented in numerical codes. This paper discusses the results of a numerical study that exam-ines the influence of adopting different rock discontinuity constitutive models for simulating the behavior of a fractured rock mass. Two constitutive approaches are employed: an enhanced Cou-lomb-based criterion with strain softening and a modified version of the Barton-Bandis model to overcome potential implementation issues. These models have been implemented in PLAXIS and their performance is inspected through numerical analyses of an underground cavity for a specific discontinuity network geometry. The results provide insights into the implications and suitability of adopting different discontinuity constitutive models for assessing the stability of engineering works in fractured rock masses.
Conference paper
(2024)
-
Hugo Portugal, Carolina Sigaran-Loria, Ronald Brinkgreve
The definition of appropriate calibration criteria for liquefaction-simulating constitutive models in sands is essential to properly simulate a liquefied state under seismic ground motions, enabling an adequate estimation of the liquefaction hazard. Therefore, a thorough quantitative characterisation of the behaviour of liquefaction is necessary to define liquefaction-triggering criteria that can describe a liquefied state in terms of stress and deformation. This paper presents a benchmark characterisation study of the behaviour of liquefaction, where liquefaction-triggering criteria, in terms of pore pressure ratio (ru) and shear strain (γ), were determined using two cyclic undrained direct simple shear test databases of Ottawa F-65 sand. The behaviour of liquefaction was analysed considering variations of initial state, confinement pressure (σ’v0) and cyclic shearing (CSR) conditions. The onset of liquefaction-triggering markers was evaluated in terms of moment of occurrence. The research concluded that liquefaction-triggering markers ru ≈ 0.95 and γ = 3% could be used simultaneously to indicate the initiation of liquefaction in loose and dense Ottawa F-65 sand under various initial and cyclic shearing conditions.
...
The definition of appropriate calibration criteria for liquefaction-simulating constitutive models in sands is essential to properly simulate a liquefied state under seismic ground motions, enabling an adequate estimation of the liquefaction hazard. Therefore, a thorough quantitative characterisation of the behaviour of liquefaction is necessary to define liquefaction-triggering criteria that can describe a liquefied state in terms of stress and deformation. This paper presents a benchmark characterisation study of the behaviour of liquefaction, where liquefaction-triggering criteria, in terms of pore pressure ratio (ru) and shear strain (γ), were determined using two cyclic undrained direct simple shear test databases of Ottawa F-65 sand. The behaviour of liquefaction was analysed considering variations of initial state, confinement pressure (σ’v0) and cyclic shearing (CSR) conditions. The onset of liquefaction-triggering markers was evaluated in terms of moment of occurrence. The research concluded that liquefaction-triggering markers ru ≈ 0.95 and γ = 3% could be used simultaneously to indicate the initiation of liquefaction in loose and dense Ottawa F-65 sand under various initial and cyclic shearing conditions.
Nowadays geotechnical engineering firms have powerful software tools to extent their consulting business also into dynamic soil-structure interaction, which before has been restricted to a rather small community of specialized experts in this field, and they certainly do. This is particularly true with respect to non-seismic sources, that is all kinds of human induced vibrations. Hence, there is a demand from clients as well as from contractors to have guidance on the requirements as well as the limits of numerical modelling of soil-structure interaction. From the literature as well as from relevant standards, recommendations for the numerical modelling of soil-structure interaction problems involving seismic actions are well known, e. g. ASCE/SEI 4-16. There are, however, some particularities when dealing with human-induced vibrations, which are absent in seismic analyses. For human-induced excitations very little specific guidance has been published in the past. A machine foundation on a homogeneous half space excited by harmonic loads with excitation frequency between 4 Hz and 64 Hz has been analysed by means of several commercially available software packages. Parametric studies have been performed to verify if recommendations for seismic soil-structure analyses are valid for non-seismic analyses as well. This paper provides details on the benchmark example and the most important conclusions from the undertaken parametric studies.
...
Nowadays geotechnical engineering firms have powerful software tools to extent their consulting business also into dynamic soil-structure interaction, which before has been restricted to a rather small community of specialized experts in this field, and they certainly do. This is particularly true with respect to non-seismic sources, that is all kinds of human induced vibrations. Hence, there is a demand from clients as well as from contractors to have guidance on the requirements as well as the limits of numerical modelling of soil-structure interaction. From the literature as well as from relevant standards, recommendations for the numerical modelling of soil-structure interaction problems involving seismic actions are well known, e. g. ASCE/SEI 4-16. There are, however, some particularities when dealing with human-induced vibrations, which are absent in seismic analyses. For human-induced excitations very little specific guidance has been published in the past. A machine foundation on a homogeneous half space excited by harmonic loads with excitation frequency between 4 Hz and 64 Hz has been analysed by means of several commercially available software packages. Parametric studies have been performed to verify if recommendations for seismic soil-structure analyses are valid for non-seismic analyses as well. This paper provides details on the benchmark example and the most important conclusions from the undertaken parametric studies.
Numerical modelling in Geo-Engineering is used to solve complex problems by simulating, analysing, or predicting soil behaviour under certain loading and boundary conditions. The soil behaviour is simulated by constitutive models that describe the relationship between stresses and strains through a mathematical formulation. Model parameters are used to calibrate model behaviour to physical soil behaviour measured during in-situ testing (e.g. CPT) or laboratory testing (e.g. triaxial testing). The selection of model parameters is challenging as it needs to cope with aspects as, constitutive model limitations, laboratory test limitations, sample disturbance, soil heterogeneity and many other. In this paper a database with over 3000 stress-strain paths measured during triaxial tests is used to derive model parameters for the Hardening Soil Small Strain Stiffness model (HS small). A procedure/algorithm has been developed to calibrate model parameters by comparing measured stress-strain paths to a simulated response from a single stress point constitutive driver. Several data analysis techniques, including machine learning tools, have been used to investigate the relationship between soil properties, soil parameters and HS small model parameters. In this paper the developed methodology and the results of the data analysis are presented.
...
Numerical modelling in Geo-Engineering is used to solve complex problems by simulating, analysing, or predicting soil behaviour under certain loading and boundary conditions. The soil behaviour is simulated by constitutive models that describe the relationship between stresses and strains through a mathematical formulation. Model parameters are used to calibrate model behaviour to physical soil behaviour measured during in-situ testing (e.g. CPT) or laboratory testing (e.g. triaxial testing). The selection of model parameters is challenging as it needs to cope with aspects as, constitutive model limitations, laboratory test limitations, sample disturbance, soil heterogeneity and many other. In this paper a database with over 3000 stress-strain paths measured during triaxial tests is used to derive model parameters for the Hardening Soil Small Strain Stiffness model (HS small). A procedure/algorithm has been developed to calibrate model parameters by comparing measured stress-strain paths to a simulated response from a single stress point constitutive driver. Several data analysis techniques, including machine learning tools, have been used to investigate the relationship between soil properties, soil parameters and HS small model parameters. In this paper the developed methodology and the results of the data analysis are presented.
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.
...
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.
Conference paper
(2024)
-
A. Laera, V.H. Miranda, R.J.N. Azeiteiro, T. Bui, S. Brasile, R.B.J. Brinkgreve
Tailings dam failures are one of the most destructive phenomena, both in terms of number of victims and of generated environmental impact. Over the years, different causes have been identified, with flow liquefaction being a prominent factor to consider when assessing the stability of tailings deposits. Due to the complexity of these events, numerical models are crucial for the analysis and design phases, where appropriate advanced soil constitutive models must be selected to reproduce the relevant features of the soil behaviour. In this paper, the Clay And Sand Model (CASM), originally proposed by Yu (1998) and later modified by Arroyo and Gens (2021) and Manica et al. (2021), has been adopted for the simulation of flow liquefaction of tailings deposits. The model incorporates the state parameter concept (Been and Jefferies, 1985) and has been implemented as a User-Defined Soil Model (UDSM) into the finite element code PLAXIS; thereby, it becomes generally applicable to a wide range of geotechnical applications. Due to its versatile yield surface and plastic potential formulations, CASM can be used to model the behaviour of a wide range of soils, from fine-grained (e.g., clays) to coarser-grained soils (e.g., silts and sands). This paper shows the capability of the model, as well as the robustness of the finite element formulation, to reproduce the soil flow liquefaction observed in boundary value problems.
...
Tailings dam failures are one of the most destructive phenomena, both in terms of number of victims and of generated environmental impact. Over the years, different causes have been identified, with flow liquefaction being a prominent factor to consider when assessing the stability of tailings deposits. Due to the complexity of these events, numerical models are crucial for the analysis and design phases, where appropriate advanced soil constitutive models must be selected to reproduce the relevant features of the soil behaviour. In this paper, the Clay And Sand Model (CASM), originally proposed by Yu (1998) and later modified by Arroyo and Gens (2021) and Manica et al. (2021), has been adopted for the simulation of flow liquefaction of tailings deposits. The model incorporates the state parameter concept (Been and Jefferies, 1985) and has been implemented as a User-Defined Soil Model (UDSM) into the finite element code PLAXIS; thereby, it becomes generally applicable to a wide range of geotechnical applications. Due to its versatile yield surface and plastic potential formulations, CASM can be used to model the behaviour of a wide range of soils, from fine-grained (e.g., clays) to coarser-grained soils (e.g., silts and sands). This paper shows the capability of the model, as well as the robustness of the finite element formulation, to reproduce the soil flow liquefaction observed in boundary value problems.
The use of Artificial Intelligence (AI) and Machine Learning (ML) have significantly increased over the last couple of years. ML is driven by the availability of data. Although geotechnical engineering is generally not among the first in picking up new technologies, there is a lot of data in our profession, and therefore, lots of opportunities to apply ML. In this article some examples are given of how ML can be used to facilitate and automate the geotechnical engineering workflow. Examples of soil identification, CPT interpretation, 3D soil stratification, parameter determination and surrogate modelling are given, and some other applications are mentioned.
...
The use of Artificial Intelligence (AI) and Machine Learning (ML) have significantly increased over the last couple of years. ML is driven by the availability of data. Although geotechnical engineering is generally not among the first in picking up new technologies, there is a lot of data in our profession, and therefore, lots of opportunities to apply ML. In this article some examples are given of how ML can be used to facilitate and automate the geotechnical engineering workflow. Examples of soil identification, CPT interpretation, 3D soil stratification, parameter determination and surrogate modelling are given, and some other applications are mentioned.
Following up on previous research on Automated Parameter Determination (APD), in which the soil stratification and numerical model parameters are automatically derived from individual CPTs, this article describes ongoing research in which the geotechnical modelling workflow is further automated in a BIM / Digital Twin environment. Especially in a preliminary project phase, when limited soil data are available, a workflow in which CPT data are used to automatically create a 3D geological model from which 2D or 3D numerical models can be extracted, may be very helpful in exploring different design alternatives. For existing (infrastructural) projects, such an automated system in a Digital Twin environment could also help responsible authorities to check the infrastructure’s safety under changing conditions. In addition to the description of technical solutions used for automatic layer detection and clustering (based on Machine Learning) across different CPTs, the article touches upon the discussion on transparency and accessibility of the automated system in view of the expertise and responsibilities of the operating geotechnical engineer.
...
Following up on previous research on Automated Parameter Determination (APD), in which the soil stratification and numerical model parameters are automatically derived from individual CPTs, this article describes ongoing research in which the geotechnical modelling workflow is further automated in a BIM / Digital Twin environment. Especially in a preliminary project phase, when limited soil data are available, a workflow in which CPT data are used to automatically create a 3D geological model from which 2D or 3D numerical models can be extracted, may be very helpful in exploring different design alternatives. For existing (infrastructural) projects, such an automated system in a Digital Twin environment could also help responsible authorities to check the infrastructure’s safety under changing conditions. In addition to the description of technical solutions used for automatic layer detection and clustering (based on Machine Learning) across different CPTs, the article touches upon the discussion on transparency and accessibility of the automated system in view of the expertise and responsibilities of the operating geotechnical engineer.
Preloading of four-legged jack-ups in clay
Geotechnical time effects and fulfilment of preloading criteria
Presented here is a numerical study on the preloading of four-legged jack-ups, such as those commonly employed in the construction of offshore wind farms. The need for reducing jack-up installation time is particularly felt within the offshore industry, especially when multiple preloading cycles are necessary in clayey soils to fulfil given preloading criteria. This is due to clays experiencing delayed deformations, causing load redistribution among all legs while the ideal situation of steady preload on all spudcans is pursued. This work employs three-dimensional finite element (3D FE) modelling to analyse the preloading performance of a reference jack-up vessel in clayey soils using a wished-in-place (WIP) approach. Detailed modelling of time effects due to soil consolidation and viscosity is introduced, with some emphasis on how to derive material parameters from typical site investigation and laboratory soil data. The results of specific parametric studies are presented to support the suitability of the adopted analysis approach, also with regard to the adoption of alternative preloading procedures. The constitutive modelling of time-dependent clay’s behaviour is shown to play a crucial role in the considered framework, and will require further research for 3D FE modelling to provide reliable quantitative support to real wind farm installation projects.
...
Presented here is a numerical study on the preloading of four-legged jack-ups, such as those commonly employed in the construction of offshore wind farms. The need for reducing jack-up installation time is particularly felt within the offshore industry, especially when multiple preloading cycles are necessary in clayey soils to fulfil given preloading criteria. This is due to clays experiencing delayed deformations, causing load redistribution among all legs while the ideal situation of steady preload on all spudcans is pursued. This work employs three-dimensional finite element (3D FE) modelling to analyse the preloading performance of a reference jack-up vessel in clayey soils using a wished-in-place (WIP) approach. Detailed modelling of time effects due to soil consolidation and viscosity is introduced, with some emphasis on how to derive material parameters from typical site investigation and laboratory soil data. The results of specific parametric studies are presented to support the suitability of the adopted analysis approach, also with regard to the adoption of alternative preloading procedures. The constitutive modelling of time-dependent clay’s behaviour is shown to play a crucial role in the considered framework, and will require further research for 3D FE modelling to provide reliable quantitative support to real wind farm installation projects.
Conference paper
(2023)
-
Tuan Anh Bui, Giuseppe Cammarata, Varun Choudary Kancharla, Ronald Brinkgreve, Sandro Brasile
Rock bolting plays an important role in different geo-engineering applications and its numerical modelling is crucial for the analysis and design of rock structures. Continuum modelling simulation of bolt-reinforced rock masses requires specific techniques to properly model the reinforcement system and its interaction with the rock mass, which often exhibits a nonlinear softening/brittle response. In this context, strain localization might occur, which, in turn, may affect numerical convergence and the quality of results. This paper presents some advanced numerical techniques implemented in PLAXIS to overcome the abovementioned challenges. Firstly, a regularization technique is implemented for an extended version of the Hoek-Brown failure criterion with strain softening. Secondly, the formulation of the structural bolt element interacting with the rock mass is developed. Finally, the robustness and accuracy of these techniques are discussed via a numerical example of a typical underground mining excavation problem.
...
Rock bolting plays an important role in different geo-engineering applications and its numerical modelling is crucial for the analysis and design of rock structures. Continuum modelling simulation of bolt-reinforced rock masses requires specific techniques to properly model the reinforcement system and its interaction with the rock mass, which often exhibits a nonlinear softening/brittle response. In this context, strain localization might occur, which, in turn, may affect numerical convergence and the quality of results. This paper presents some advanced numerical techniques implemented in PLAXIS to overcome the abovementioned challenges. Firstly, a regularization technique is implemented for an extended version of the Hoek-Brown failure criterion with strain softening. Secondly, the formulation of the structural bolt element interacting with the rock mass is developed. Finally, the robustness and accuracy of these techniques are discussed via a numerical example of a typical underground mining excavation problem.
An ongoing research project aims to create an automated parameter determination (APD) framework relying on a graph-based approach for determining constitutive model parameters from in-situ tests. The system requires two spreadsheets as inputs. One spreadsheet defines the parameters, while the other spreadsheet specifies the correlations. The system connects parameters and methods by generating paths between them and calculates the value(s) for different parameters. So far, the frame-work focused on determining soil parameters based on the cone penetration test (CPT). This paper focuses on expanding the framework by adding the dilatometer test (DMT). A new database of correlations for the DMT is compiled. The expanded APD framework successfully calculates soil parameters for coarse and fine-grained soils based on CPT as well as DMT data. Validat-ing the output of the system, assessing the accuracy of the derived parameters, and connecting soil parameters to constitutive model parameters are part of ongoing research.
...
An ongoing research project aims to create an automated parameter determination (APD) framework relying on a graph-based approach for determining constitutive model parameters from in-situ tests. The system requires two spreadsheets as inputs. One spreadsheet defines the parameters, while the other spreadsheet specifies the correlations. The system connects parameters and methods by generating paths between them and calculates the value(s) for different parameters. So far, the frame-work focused on determining soil parameters based on the cone penetration test (CPT). This paper focuses on expanding the framework by adding the dilatometer test (DMT). A new database of correlations for the DMT is compiled. The expanded APD framework successfully calculates soil parameters for coarse and fine-grained soils based on CPT as well as DMT data. Validat-ing the output of the system, assessing the accuracy of the derived parameters, and connecting soil parameters to constitutive model parameters are part of ongoing research.
Selecting an appropriate soil constitutive model and determining the corresponding model parameters for numerical analysis are considered most challenging in geotechnical engineering. While many empirical relationships have been proposed to derive soil parameters from in situ test results, there is no clear procedure on how to derive model parameters uniquely. In practice, available data during the early stages of projects is often limited to field test data. Consequently, different engineers provide different numerical solutions for the same problem. As a solution, the authors present a proof of concept for an automated parameter determination (APD) system, using concepts of graph theory to determine constitutive model parameters from in situ tests while keeping the system transparent (verifiable) and adaptable (extendable). The study aims to increase the confidence in parameter determination for numerical analysis by giving the user of the system, the geotechnical engineer, control over the system. Using a spreadsheet of parameters and equations as input, the system generates paths between the parameters and calculates the parameter values for coarsegrained soil, starting from CPT data. Further validation and tweaking of the system, as well as the extension to other types of soils, are part of ongoing research. ...
Selecting an appropriate soil constitutive model and determining the corresponding model parameters for numerical analysis are considered most challenging in geotechnical engineering. While many empirical relationships have been proposed to derive soil parameters from in situ test results, there is no clear procedure on how to derive model parameters uniquely. In practice, available data during the early stages of projects is often limited to field test data. Consequently, different engineers provide different numerical solutions for the same problem. As a solution, the authors present a proof of concept for an automated parameter determination (APD) system, using concepts of graph theory to determine constitutive model parameters from in situ tests while keeping the system transparent (verifiable) and adaptable (extendable). The study aims to increase the confidence in parameter determination for numerical analysis by giving the user of the system, the geotechnical engineer, control over the system. Using a spreadsheet of parameters and equations as input, the system generates paths between the parameters and calculates the parameter values for coarsegrained soil, starting from CPT data. Further validation and tweaking of the system, as well as the extension to other types of soils, are part of ongoing research.