Open Master’s projects at LMS

Context

• Seismic analysis of existing buildings
• Role of soil-structure interaction (SSI) often neglected

Objectives

• How should SSI be accounted for to improve reliability of seismic vulnerability assessments?
• Inform appropriate retrofit strategies

Activities

• Synthesize current research and regulatory frameworks regarding SSI effects to identify critical cases
• Use numerical modelling approaches that integrate SSI effects
• Apply developed methodology to a case study
• Report the results in a structured form

Contact: Dr. Aldo Madaschi ([email protected])

Context

• Borehole heat exchangers (BHEs)
• Machine learning approach

Objectives

• Predict heat transfer performance of BHEs using machine learning techniques

Activities

• Familiarise with COMSOL Multiphysics
• Understanding of the current formulation of numerical models for BHEs
• Identification and development of machine learning technique for BHEs
• Report the results in a structured form

Supporting documentation

Contact: Dr. Elena Ravera ([email protected])

Context

• Borehole heat exchangers (BHEs)
• Machine learning approach

Objectives

• Data collection from instrumented pilot installations
• Predict heat transfer performance of BHEs using machine learning techniques

Activities

• Installation of non-intrusive temperature sensors on hydraulic loops
• Verification of the quality and reliability of the measurements
• Identification and development of machine learning technique for BHEs
• Report the results in a structured form

Supporting documentation

Contact: Dr. Elena Ravera ([email protected])

Context

• The caprock layer (e.g. shale) prevents injected CO₂ from escaping to the surface
• 3D micro-structural characterisation with x-ray tomography

Objectives

• Understand sealing capacity of fissured and crushed caprock material

Activities

• Familiarise with basic image analysis and processing (Fiji ImageJ)
• Characterise caprock micro-structural heterogeneity
• Quantify volumetric changes due to hydraulic and mechanical changes (digital volume correlation)
• Report the results in a structured form

Contact: Dr. Eleni Stavropoulou ([email protected])

Context

• Geological CO₂ storage
• Machine learning modelling

Objectives

• Synthetic data generation
• Predict hydromechanical response based on machine learning techniques

Activities

• Familiarise with FEM CodeBRIGHT
• Produce a large dataset of synthetic hydromechanical data
• Identification and development of data-driven machine learning models for geological CO₂ storage
• Report the results in a structured form

Contact: Dr. Eleni Stavropoulou ([email protected])

Context

• Underground engineering decisions strongly depend on reliable knowledge of the in situ stress state
• Today, the full stress state cannot be determined with sufficient confidence from borehole measurements under realistic geological conditions
• The IS-E experiment at Mont Terri provides a unique opportunity to address this limitation through a dedicated multi-borehole stress-measurement campaign
• The dataset will support the development and validation of advanced interpretation frameworks for full stress-state reconstruction in borehole environments

Objectives

• Reconstruct and interpret in situ stress components from field measurements
• Validate the EPFL-LMS methodology for full stress-state reconstruction in borehole environments

Activities

• Familiarisation with the Mont Terri experiment and stress-measurement concept
• Data cleaning, organisation and quality checks of field measurements
• Analytical and computational post-processing of the stress dataset
• Uncertainty assessment and comparison across test locations
• Reporting and presentation of results in a structured form

Contact: Dr. Angelica Tuttolomondo ([email protected])

Context

• The hydro-mechanical behaviour of unsaturated soils is governed by complex water–soil interactions
• Water is retained through different mechanisms, including adsorption at the particle scale and capillarity within the pore network
• These mechanisms contribute differently to the mechanical behaviour of soils; distinguishing them is therefore essential for a better interpretation of water retention and deformation response

Objectives

• Provide new insights into soil water retention mechanisms across different conditions
• Quantify the contribution of adsorbed and capillary water to the overall retention behaviour

Activities

• Literature review on water retention mechanisms and testing approaches
• Experimental testing on prepared soil samples
• Data processing and critical discussion of the results
• Collaboration with the University of Genoa and ENTPE Lyon
• Reporting and presentation of results in a structured form

Contact: Dr. Angelica Tuttolomondo ([email protected])

Context

• Geotechnical infrastructure design requires reliable subsurface models
• Cone Penetration Tests (CPTs) widely used for stratigraphic characterisation
• Increasing availability of large geotechnical databases
• AI/machine learning approaches for automated soil layer identification

Objectives

• Develop AI-based workflows for automated stratigraphic identification
• Quantify uncertainty in layer boundary identification
• Support scalable 3D subsurface modelling for infrastructure projects

Activities

• Familiarisation with CPT and borehole databases
• Data preprocessing and cleaning of large geotechnical datasets
• Development and testing of AI/ML algorithms for layer identification
• Interpretation of uncertainty and engineering implications
• Reporting and presentation of results in a structured form

Contact: Dr. Angelica Tuttolomondo ([email protected])