Manuel López Aenlle

A general procedure for estimating dynamic displacements using strain measurements and operational modal analysis

Measurement systems are being installed in more and more civil structures with the purpose of monitoring the general dynamic behavior of the structure. The instrumentation is typically done with accelerometers, where experimental frequencies and mode shapes can be identified using modal analysis and used in health monitoring algorithms. But the use of accelerometers is not suitable for all structures. Structures like wind turbine blades and wings on airplanes can be exposed to lightning, which can cause the measurement systems to fail. Structures like these are often equipped with fiber sensors measuring the in-plane deformation. This paper proposes a method in which the displacement mode shapes and responses can be predicted using only strain measurements. The method relies on the newly discovered principle of local correspondence, which states that each experimental mode can be expressed as a unique subset of finite element modes. In this paper the technique is further developed to predict the mode shapes in different states of the structure. Once an estimate of the modes is found, responses can be predicted using the superposition of the modal coordinates weighted by the mode shapes. The method is validated with experimental tests on a scaled model of a two-span bridge installed with strain gauges. Random load was applied to simulate a civil structure under operating condition, and strain mode shapes were identified using operational modal analysis.

Experimental Evaluation of Mass Change Approaches for Scaling Factors Estimation

In operational modal analysis the input forces are unknown so the modes shapes are obtained with an arbitrary normalization. In some applications the mass normalized modes shapes are required and the scaling factors have to be estimated. A way of obtaining these factors is to modify the modal properties of the structure by introducing changes in the mass or stiffness; mass change is usually the most convenient way to modify the structural properties in a controlled manner. Different strategies and approaches to perform the mass change method have been proposed in recent years. In this work the scaling factors of a 15 tonnes concrete slab strip are obtained using several equations proposed by different authors and the results are compared. The results show that the scaling factors can be estimated with reasonable accuracy using the different equations if a good mass change strategy is applied.

Strain Estimation in a Glass Beam Using Operational Modal Analysis

A potential application of operational modal analysis is the prediction of strain or stress time histories which, on the other hand, are one of the most important sources of uncertainty in fatigue design and remaining fatigue life calculations. This is due to the difficulty of estimating the stiffness, mass and damping properties with accuracy, as well as the use of simplified loading models.

Detection of Mass Change on a Glass Plate

In model updating the aim is to reach the optimal correlation between FE model and test data by modifying model parameters. The traditional solution to an updating problem is obtained by non-linear optimization processes governed by the selected updating parameters and target responses. Due to imprecision and lack of information in measurements, inaccuracy in model and several possible updating parameters a wide range of potential solutions to the same problem is present. This paper proposes a technique in which the updating problem is solved in one step using modal properties, i.e. natural frequencies and mode shapes, as target responses. The method utilizes mode shape sensitivity equations combined with the Bernal Projection equation to detect mass and stiffness discrepancies between model and experimental data. The proposed one step solution will only work accurately in cases where a reasonable FE model is available. The technique is demonstrated on simulated data of modal properties before and after mass perturbation of a glass plate. The data is polluted with noise in the range of what can be expected from real measured data.

Fatigue damage detection and prediction of fatigue life on a cantilever beam

Purpose Fatigue failure is an important criterion to be considered in the design of structures and mechanical components. Catastrophic failure of structures in service conditions can be avoided using adequate techniques to detect and localize fatigue damage. Modal analysis is a tool used in mechanical and structural engineering to estimate dynamic properties and also to monitor the health of structures. If modal analysis is applied periodically to a structure, fatigue damage can be detected and localized and the fatigue life can be extended by means of suitable reinforcement and repairing. The paper aims to discuss these issues. Design/methodology/approach The experimental results corresponding to the fatigue tests carried out on a steel S-275 cantilever beam are presented. Operational modal analysis was applied periodically to the beam in order to study the variation of modal parameters during the tests and the stresses were estimated combining a numerical model and the acceleration modal coordinates measured at discrete points of the structure. The experimental results are compared with those predicted applying the S-N model of Eurocode 3. Findings A methodology that combines a finite element model and the experimental responses of a structure has been applied to estimate the stress time histories of a cantilever beam clamped to a foundation through a steel plate. The estimated stresses have been used to predict the fatigue damage according to the Eurocode 3. Due to the fact that no information of the scatter is provided by this code (EC3), only the number of cycles corresponding to a probability of failure of 5 percent can be predicted. Originality/value The proposed methodology can be applied to real structures in order to know the accumulated fatigue damage in real time.

Cross Orthogonality Check for Structures with Closely Spaced Modes

Structures with closely spaced modes can often be hard to correlate with numerical models due to the high sensitivity of the eigenvectors. Even the smallest change in either mass or stiffness can have a large influence on the eigenvectors, and makes it hard to fit a numerical model so its modal parameters matches those obtain from measurements. This paper introduces a robust method for calculating the cross orthogonality check for structures with closely spaced modes. The method utilizes the fact that a cluster of closely spaced eigenvectors from an experiment and from a well correlated numerical model will span the same subspace, although the experimental mode is badly correlated with its corresponding numerical mode. A new basis of numerical modes is created by redefining the closely spaced numerical modes as a linear combination of one another, based on their projection upon the experimental mode. This will enable a more robust calculation of the cross orthogonality check. The method is validated using simulation cases where the errors are evaluated using simulated responses for the different sets of modal parameters.

Dynamic Behavior of Laminated Glass Beams

Laminated glass beams can be considered sandwich elements consisting of two or more glass sheets and one or more interlayers such as polyvinyl butyral (PVB). The response of these elements present two borderline cases: (1) the monolithic limit, which appears at low temperatures and high frequencies and (2) the layered limit, which appears at high temperatures and low frequencies. In this work, the modal parameters of several laminated glass beams were estimated by operational modal analysis in the temperature range 10–45 °C. The results show that the natural frequencies decrease and the damping ratios increase with increasing temperature whereas there are not significant changes in the mode shapes. The experimental results are compared with those determined using the analytical models of Mead and Markus and Ross, Kerwin and Ungar.

Influence of the Support Conditions in the Modal Parameters of a Cantilever Beam

In modal analysis, previously to the modal testing, it is very common to assembly a finite element model, in order to know the frequency range of interest and the mode shapes of a structure. This information is very helpful for planning and developing the modal testing.