Master Theses

Model Reduction of Multibody Dynamics in Railways

Supervisors: Dr. Charikleia Stoura and Prof. Dr. Eleni Chatzi (ETH Zürich), Prof. Shobhit Jain (TU Delft),

Content
Multibody models are widely used to simulate railway vehicle dynamics, including the interactions between car bodies, bogies, wheelsets, suspensions, and wheel–rail contact. Despite their relatively limited number of degrees of freedom, repeated simulations can become computationally demanding, particularly for parametric studies, uncertainty quantification, optimization, and real-time applications.

This Master’s thesis aims to investigate and implement model-order-reduction techniques for rigid multibody railway vehicle systems. The objective is to develop computationally efficient reduced-order models (ROMs) that accurately reproduce the dominant vehicle dynamics while retaining the influence of suspension and wheel–rail contact forces across different operating conditions.

Expected project outcomes:

• Review model reduction techniques for rigid multibody dynamical systems.
• Develop a representative railway vehicle model and validate it against specialized software such as SIMPACK.
• Implement suitable reduction techniques to construct reduced-order railway vehicle models.
• Evaluate model accuracy, stability, and computational efficiency against the full-order model.
• Assess robustness under varying vehicle speeds, track irregularities, loading conditions, and suspension parameters.

Suggested Competencies: Linear algebra, numerical methods, MATLAB or Python programming. Background in structural dynamics, multibody systems, or model reduction is beneficial

Collaboration: Delft Institute of Applied Mathematics, TU Delft, and Chair of Structural Mechanics and Monitoring, ETH Zürich.

Graph-Based Reinforcement Learning for Erosion-Safe Operation and Maintenance of Offshore Wind Farms

Supervisors:  Lidiya Kalathiparambil Kennedy (Chair of Structural Mechanics and Monitoring, ETH Zürich), Alessandro Bassalti (Siemens Digital Industries Software, Leuven, and CONSTRUCT, Faculty of Engineering, University of Porto), Prof. Dr. Eleni Chatzi (Chair of Structural Mechanics and Monitoring, ETH Zürich).

Content
Leading Edge Erosion (LEE), caused by rain and hail impacts on wind turbine blades, reduces energy production and drives costly offshore repairs. Erosion-safe operation (ESO), which means temporarily reducing blade tip speed during adverse weather, can extend blade lifetime at a small cost in energy. Reinforcement learning (RL) has shown promise for learning such strategies for single turbines, but in a wind farm, turbines are coupled through wake effects and shared maintenance logistics.

This thesis develops a graph-based RL framework in which the wind farm is represented as a graph (turbines as nodes, wake interactions as edges). The agent decides how much to derate each turbine during adverse weather and when to dispatch maintenance, balancing energy revenue against erosion damage, repair costs, and the effect of derating on structural fatigue loads.

Expected project outcomes

1. Review literature on leading edge erosion, erosion-safe operation, and graph neural networks for control
2. Extend an existing wind farm simulator (wake effects, fatigue) with erosion progression, aerodynamic degradation for maintenance conditioned erosion safe operations.
3. Define a graph representation of the wind farm suitable for RL controlTrain the agent and compare the learned policy against baseline strategies

Suggested Courses/Background: Structural Dynamics, Structural Identification & Health Monitoring, a course on machine learning, Aerodynamics, Wind Energ

Suggested Competencies: Proficiency in Python; familiarity with PyTorch or willingness to learn; interest in wind energy and structural reliability

Digital Twins and Temperature-Aware Monitoring for Bridge Performance Assessment

Supervisors: Dr. Yves Reuland (irmos technologies AG), Dr. Minas Spyridonakos (irmos technologies AG), Prof. Dr. Eleni Chatzi (ETH Zürich)

Content
Bridges are critical components of modern transport networks and must remain safe and functional well beyond their original design life. Increasing traffic loads, environmental variability, and aging materials challenge traditional assessment approaches based on periodic inspections and conservative safety margins.

Continuous monitoring combined with physics-based and data-driven modeling enables the development of digital twins — dynamic representations of structural behavior under real operating conditions. Such frameworks support safety assessment, help define thresholds for normal behavior, and improve early damage detection.

A central challenge lies in understanding the interaction between operational loads (traffic, wind) and environmental influences, particularly temperature, which induces spatially and temporally varying structural responses. This thesis offers two complementary research directions within this framework.

Possible Direction A: Data-Driven Digital Twin Development

Focus: Understanding how operational and environmental conditions influence structural dynamics and embedding this knowledge into predictive models.

Main tasks:

• Analyze real monitoring data to quantify the influence of traffic and environmental conditions on modal properties and structural response.
• Calibrate and validate a physics-based bridge model using measured data.
• Develop a reduced or surrogate model to predict structural behavior under varying operational scenarios.
• Define data-informed indicators and thresholds for SHM applications.

Possible Direction B: Temperature-Induced Structural Dynamics

Focus: Explicit modeling of temperature effects and their impact on monitoring interpretation

Main tasks:

• Develop a Finite Element model incorporating temperature-dependent material behavior.
• Simulate spatially varying temperature fields and evaluate their effect on modal characteristics.
• Compare simulated temperature-induced dynamics with monitoring data.
• Assess how temperature variations may mask or mimic damage-related signatures.
  

Suggested Courses: Structural Identification and Health Monitoring, course on machine learning methods

Suggested Competencies: Python/MATLAB

Improving the design of rotating parts of wind turbines through metamaterial technology

Supervisors: Dr. Vasilis Dertimanis (), Prof. Dr. Eleni Chatzi

Industry Partners: Dr. Luca Sangiuliano, Phononic Vibes

Content
In the context of novel lightweight materials for noise and vibration reduction, Phononic Vibes Srl (https://phononicvibes.com/) is researching, designing and producing metamaterial technology. Metamaterials are often periodic structures, consisting of a host structure with embedded or added resonant structures, leading to superior vibro-acoustic attenuation performance in some tunable frequency ranges, called stop bands. In the following videos, the metamaterial concept is applied structure-borne (https://www.youtube.com/watch?v=DUee93HcPVQ&t) and airborne noise problems.

The resonant structures often have some small features that control their tuned frequency and the dynamic behavior of the entire metamaterial system. Once these resonant structures are applied to rotating systems, the stop band behaviour can be modulated according to the rotation speed to the entire system thanks to the presence of inertial forces (ref. 1). This interesting phenomenon can give rise to new methods to control the dynamic behavior of rotating systems and to reduce the unwanted dynamics.

Rotating parts of wind-turbines are often subjected to extreme dynamic conditions. To ensure the long service life and optimal design of rotating parts of wind-turbines, thus preventing failures and expensive maintenances, the dynamic behaviour of these parts must be thoroughly designed and controlled.

The goal of this master thesis is then to investigate the potential of metamaterial technology applied to the rotating parts of wind turbines to control and preserve their dynamic behaviour. In particular, the student will explore how to exploit the modulation effect given by the inertial forces to tune the metamaterial system to reduce critical dynamic conditions. Analytical and numerical models will be developed by the student to study the physics and to assess the metamaterial performance applied to such systems. The aim is to propose design guidelines to achieve a robust design solution. 

This thesis is in collaboration with Phononic Vibes Srl and is mainly simulation-based.

Suggested Competencies: Python/MATLAB

Vibration Monitoring and Weigh-in-Motion of an In-Service
Road Bridge with Fiber Segment Interferometry (FSI)

Supervisors: Dr. Charikleia Stoura, Prof. Dr. Eleni Chatzi 

Industry Supervisor: Filippo Rossi, KISTLER Instrumente AG

Content 
Fiber Segment Interferometry (FSI) measures strain along multiple segments of an optical fiber. Combining high resolution with sampling rates up to 410 kHz, it can capture quasi-static traffic loads, structural vibrations and acoustic emission using the same sensor. Previous experiments on the openLAB research bridge demonstrated its potential for detecting and localizing tendon-wire breaks and identifying structural vibration modes. This Master’s thesis will investigate FSI for vibration-based structural monitoring and bridge weigh-in-motion on an in-service road bridge in Winterthur. The bridge is equipped with four FSI sensors arranged longitudinally and transversely, alongside an existing monitoring system comprising acceleration, inclination and strain sensors. Working with Kistler engineers, the student will plan and conduct measurement campaigns covering ambient and traffic-induced vibrations, crossings by vehicles of known weight, and controlled impacts. The objective is to assess how FSI can identify bridge dynamics and estimate vehicle loads under operational conditions. A simplified finite-element model will support interpretation of the measurements and weigh-in-motion analysis.

Expected Project Outcomes

• Characterize FSI signals, including traffic events, noise levels and recurring vibrations.
• Apply operational modal analysis to identify natural frequencies, damping and strain mode shapes, and compare results with existing bridge sensors.
• Develop and calibrate weigh-in-motion methods for axle detection, vehicle speed and gross vehicle weight estimation.
• Survey the bridge geometry and develop a simplified finite-element model, updated using the identified modes.
• Assess temperature and traffic effects on modal parameters to support long-term monitoring.
• Conduct an exploratory study using Schmidt hammer impacts to assess stress-wave detection, wave speed and impact localization.

Suggested courses: Structural Dynamics and Vibration Problems; Structural Identification & Health Monitoring; FEM I/II 

Suggested competencies: Interest in field work and signal processing; basic knowledge of structural dynamics and modal analysis

Software knowledge: Python (NumPy, SciPy; pyOMA or similar) or MATLAB; an FE tool (e.g. OpenSees, Abaqus, ANSYS or SAP2000)