Progress in Computational Structures Technology
Edited by: B.H.V. Topping and C.A. Mota Soares

Chapter 7

Identification and Control of Structural Systems

O.S. Bursi*, L. Vulcan*, W. Salvatore+ and L. Nardini+
*Department of Mechanical and Structural Engineering, University of Trento, Italy
+Department of Structural Engineering, University of Pisa, Italy

Keywords: structural system identification, health monitoring, curvature and displacement mode shape, adaptive control, model reference, stability analysis.

The modelling of complex structures subjected to earthquake or dynamic loading sometimes experimentally validated as well as active control has been a subject of intense research in the last decades. Due to the governing assumptions of linearity in the analysis, to the statistical variability in the properties of structural components owing to fabrication tolerance, and to the increased heterogeneity of the structures being modelled, a systematic improvement in the reliability of finite element (FE) analysis is difficult to achieve. Therefore, test-validated FE models still play a key role in the design of high reliable structures. Modal testing is perhaps the most versatile form of structural model validation testing. Nonetheless, modal testing is not a straightforward technique as the verification of the dynamic performance of a structure remains strongly dependent on the choice of analytical models. Then, unlike mass property measurements, modal testing captures global response measures which characterize mass, damping and stiffness behaviour, simultaneously. As a result, the use of multiple simultaneous random excitations with multiple output readings, arguably the multiple input/output testing, is the most important modal testing method used today.

An important area of application of modal testing based on system identification is structural health monitoring, in view of detection, location and quantification of damage. This practice has taken on increased importance in civil applications owing to the increased use of structures far beyond their original life expectancy, which underlies the concept of open-loop smart structures. In detail, vibration-based health monitoring algorithms can be classified into two categories: i) model-based approaches; ii) non-model-based approaches. Model-based techniques use changes in response functions or modal parameters such as natural frequencies, mode shapes, or their derivatives, in order to identify damage locations and levels. Analysis of changes of parameters between sequential tests over time is used to determine damage characteristics. Some of the algorithmic approaches include mode shape-base techniques and flexibility methods. Conversely, non-model-based schemes determine direct changes in the sensor output signal in order to locate damage in the structure. These can be thought of as signal-processing solutions to the problem.

Model-based damage detection methods can also be classified with regard to global or local response characteristics. In a greater detail, the changes in a structure modal parameters could be detected either using the global modes directly, or using subassemblies of the structure. This localised approach avoids to consider the entire system provided that enough measurements to characterize the subassembly under consideration are available.

Some novel measuring technologies of interests are Fibre Optic Sensors (FOSs) and Micro-Electro-Mechanical Systems (MEMSs). Many applications of FOSs have been recently proposed in the field of structural engineering, in order to measure quantities such as strains, temperatures, moisture content. The possibility to embed FOSs into the structure during the construction process allows one to monitor the strains or other parameters at critical locations where high values are expected. These advantages, together with excellent durability properties, make the FOS an effective tool in the lifelong monitoring of structural elements and structures as a whole. With regard to FOSs, many applications of Fibre Bragg Grating sensors are reported in the monitoring of bridges and viaducts.

A key evolution in sensors is toward miniaturisation employing MEMSs. The sensor itself consists of a thin silicon foil, even if commonly comes mounted in the form of a micro-chip. Today, a number of MEMS sensors is available off-the-shelf, including transducers accelerometers, pressure gauges, load cells, gyroscopes and chemical gauges. Moreover, they can be wireless, i.e. they need power, but no cables for signals. MEMSs application in civil engineering is theoretically feasible, even though few applications can be found in the literature.

The subject of structural control offers opportunities to design new structures and to retrofit existing ones by the application of counter-forces, smart materials, frictional devices, etc. instead of just increasing the strength of the structure at greater cost. A variety of applications has already been installed in building structures to control wind induced motions that are objectionable to the occupants; and many applications of passive control, such as base isolations have been installed to reduce structural acceleration produced by strong earthquake ground shaking. There is a consensus that structural control has the potential for improving the performance of structures, new or existing, if appropriate research and experimentation are undertaken.

To begin, this chapter introduces concepts relevant to smart structures and an example of an open-loop smart element that can be employed in structural engineering. Then, two applications of model-based structural health detection techniques are shown. In detail, the use of FOSs and classical accelerometers in conjunction with curvature mode shapes and localised damage techniques are used, respectively, for damage detection and quantification. Successively, the theory of adaptive control is illustrated and a real-time compatible algorithm based on a Rosenbrock method is introduced, for an adaptive minimal control synthesis algorithm formulated in discrete form, in view of a stability analysis of a non-linear controller. Finally, conclusions are drawn with some remarks on future developments.

return to the contents page