Álvaro Cunha

The reliable identification of a structure’s modal properties from in situ vibration data often uses parametric, multi-order methods like time-domain SSI-Cov or frequency-domain p-LSCF (PolyMAX). Traditionally, this process relies on manual interpretation of complex stabilisation diagrams, which is time-consuming and prevents efficient, long-term continuous monitoring. To overcome this, the field has successfully shifted toward Automated Operational Modal Analysis (AOMA), leveraging clustering algorithms to automatically group similar modal estimates obtained across several model orders. However, in the case of dam monitoring, though successful results have been achieved with hierarchical clustering, a significant challenge persists: harmonic excitation, particularly from nearby power plants, can introduce strong non-structural frequencies that are easily mistaken for true vibration modes, creating a source of bias in the identified modal properties. To improve the robustness of automated systems, it is essential to develop methodologies that reliably distinguish between legitimate structural estimates and these external “intruders”. This work proposes a post-tracking methodology based on Gaussian Mixture Models applied to tracked modal estimates of a concrete arch dam. By combining features derived from identified poles uncertainties, damping ratios, and frequency stationarity, the method separates structural modal estimates from harmonic contamination without relying on site-specific thresholds, exploiting the statistical contrast between diffuse structural populations and highly concentrated harmonic signatures.

This work focuses on the dynamic monitoring behaviour of concrete dams, with a specific emphasis on the Baixo Sabor dam as a case study. The main objective of the dynamic monitoring is to continuously observe the dam’s behaviour, ensuring it remains within expected patterns and issuing alerts if deviations occur. The monitoring process relies on on-site instruments and behaviour models that use pattern recognition, thereby avoiding explicit dependence on mechanical principles. The undertaken work aimed to develop, calibrate, and compare statistical and machine learning models to aid in interpreting the observed dynamic behaviour of a concrete dam. The methodology included several key steps: operational modal analysis of acceleration time series, characterisation of the temporal evolution of observed magnitudes and influential environmental and operational variables, construction and calibration of predictive models using both statistical and machine learning methods, and the comparison of their effectiveness. Both Multiple Linear Regression (MLR) and Multilayer Perceptron Neural Network (MLP-NN) models were developed and tested. This work emphasised the development of several MLP-NN architectures. MLP-NN models with one and two hidden layers, and with one or more outputs in the output layer, were performed. The aim of this work is to assess the performance of MLP-NN models with different numbers of units in the output layer, in order to understand the advantages and disadvantages of having multiple models that characterise the observed behaviour of a single quantity or a single MLP-NN model that simultaneously learns and characterises the observed behaviour for multiple quantities. The results showed that while both MLR and MLP-NN models effectively captured and predicted the dam’s behaviour, the neural network slightly outperformed the regression model in prediction accuracy. However, the linear regression model is easier to interpret. In conclusion, both methods of linear regression and neural network models are suitable for the analysis and interpretation of monitored dynamic behaviour, but there are advantages in adopting a single model that considers all quantities simultaneously. For large-scale projects like the Baixo Sabor dam, Multilayer Perceptron Neural Networks offer significant advantages in handling intricate data relationships, thus providing better insights into the dam’s dynamic behaviour.

Operational Modal Analysis is a reliable methodology for the assessment of civil engineering structures, allowing for the accurate definition of their dynamic behavior. Additionally, since it does not require the use of artificial excitation, it becomes a cost-effective choice for the performance of singular tests, as well as a consistent option for the long-term continuous monitoring of structures. Nevertheless, the modal characteristics of structures are affected by environmental and operational conditions, concealing the variations that could emerge due to abnormal behavior. With respect to concrete dams, factors such as temperature and the level of water in the reservoir exert a pronounced influence in the evolution of natural frequencies, increasing data variability and camouflaging the behavior that would be discerned under stable conditions. In this context, the current study seeks to examine the ability of statistical and machine learning tools to mitigate the effects of external conditions on the modal properties of concrete dams, specifically natural frequencies. To achieve this objective, the efficiency of methods incorporating measurements of variables impacting the structure, such as Multiple Linear Regressions and Neural Networks, is compared to that of tools not needing these inputs, as is the case of Principal Components Analysis and the Minimum Mean Square Error estimator. Experimental data obtained during the continuous dynamic monitoring of a concrete dam in Portugal is used as a case study.

Due to the increasing importance of the continuous monitoring of Civil Structures, this research aims to take advantage of new solutions of measurement systems and sensors in the Structural Health Monitoring of bridges, using the reinforced concrete arch Arrábida Bridge as a case study. With the support of customized sensors, this work starts by performing preliminary ambient vibration tests on Arrábida Bridge, aiming at the identification of the natural frequencies and respective vibration modes of the deck. Then, the measurement campaigns carried over time are described, which involved different types of customized sensors, namely, accelerometers, temperature sensors and displacement sensors. Based on the signals collected by these devices, some preliminary analyses were performed. The results show that the temperature measured at the deck sections presents different amplitudes and phase shifts when compared to the temperature measured at the arch sections. Moreover, using the temperature measurements, it is possible to estimate with good accuracy the displacements in the expansion joints of the bridge. It was also observed that the displacements in these joints, although being conditioned by the temperature effects, are also marked by a dynamic component arising from the traffic loads over the deck. The observation of this phenomenon is an innovative aspect found in this investigation, which can be used in the future to characterize the traffic loads on the structure.

Wind turbines are structures predominantly subjected to dynamic loads throughout their period of life. In that sense, fatigue design plays a central role. Particularly, support structure design might be conservative with respect to fatigue, which may lead to a real fatigue life of considerably more than 20 years. For these reasons, the implementation of a fatigue monitoring system can be an important advantage for the management of wind farms, providing the following outputs: (i) estimation of the evolution of real fatigue condition; (ii) since the real condition of fatigue damage is known, these results could be an essential element for a decision about extending the lifespan of the structure and the possibility of repowering or overpowering; and (iii) the results of the instrumented wind turbines can be extrapolated to other wind turbines of the same wind farm. This paper reviews the procedures for calculating the fatigue damage of wind turbine towers using strain measurements. The applicability of the described procedures is demonstrated with experimental data acquired in an extensive experimental campaign developed at Tocha Wind Farm, an onshore wind farm located in Portugal, exploring the impact of several user-defined parameters on the fatigue results. The paper also includes the description of the data processing needed to convert raw measurements into bending moments and several validation and calibration steps.