Structural Damage Assessment using an Artificial Immune System

Structural Damage Assessment using an Artificial Immune System

Maribel Anaya Vejar (Universitat Politècnica de Catalunya, Spain & Universidad Santo Tomás, Colombia), Diego Alexander Tibaduiza Burgos (Universidad Santo Tomás, Colombia) and Francesc Pozo (Universitat Politècnica de Catalunya, Spain)
Copyright: © 2015 |Pages: 17
DOI: 10.4018/978-1-4666-8490-4.ch005
OnDemand PDF Download:


Structural damage assessment methodologies allow providing knowledge about the current state of the structure. This information is important because allows to avoid possible accidents and perform maintenance tasks in the structure. This chapter proposes the use of an artificial immune system to detect and classify damages in structures by using data from a multi-actuator piezoelectric system that is working in several actuation phases. In a first step of the methodology, principal component analysis (PCA) is used to build a baseline model by using the collected data. In a second step, the same experiments under similar conditions are performed with the structure in different states (damaged or not). These data are projected into the different baseline models for each actuator, in order to obtain the damages indices and build the antigens. The antigens are compared with the antibodies by using an affinity function and the result of this process allows detecting and classifying damages.
Chapter Preview


Structural health monitoring (SHM) is an important discipline which seeks to assess the proper performance of a structure. To achieve this objective, SHM makes use of sensors permanently installed in the structure for inspecting and defining its current state based on the analysis of structural responses. As a result, the collected structural responses are analyzed and compared with baseline patterns in order to detect abnormal characteristics and define the structural integrity. The obtained information can be used to define whether the structure can operate and under which conditions.

In general, the damage identification can be performed by two main approaches; the first consists in obtaining a reliable physics-based model of the structure, while the second is based on data-driven approaches which normally tackle the problem as one of pattern recognition.

One advantage of the use of data driven approaches is the reliability in the analysis since the indication of damage could be directly determined with the comparison between a baseline and the data collected. However, to ensure the reliability of the analysis performed to the signals collected in several experiments, it is necessary to ensure the proper functioning of the sensors, actuators and hardware used to inspect the structure. Among the big quantity of damages that can be presented in the normal service of a structure, the following categories can be distinguished (Farrar, Doebling, & Nix, 2001), (Doebling, Farrar, & Prime, 1998):

  • Gradual damage such as fatigue, corrosion, and aging.

  • Sudden and predictable damage like aircraft landing and planned explosions in confinement vessels.

  • Sudden and unpredictable damage originating from foreign-object-impact, earthquakes and wind loads.

At the same time, these different kinds of damage can also be classified depending on its severity in three big groups (Tibaduiza, 2013):

  • Light Damage: This corresponds to the initial stage of damage, which can be relatively easily-repairable and is not dangerous for the normal operation of the structure.

  • Moderate Damage: In comparison with the previous one, this damage requires major repairs and need to be evaluated more carefully in order to define if the structure can operate in normal conditions.

  • Severe Damage: This type of damage unlike previous damages requires big reparations or the replacement of the structure.

For a structure in service, the variability in its dynamic properties can be a result of time-varying in environmental and operational conditions. Some of the environmental conditions to consider are humidity, wind loads, temperature, and pressure, among others. Operational conditions include loading conditions, operational speed and mass loading (Sohn, 2007). The damage identification techniques need in this way to consider several variables and most of the cases the damage identification procedure depends on the critical damage admissible in the structure.

The damage diagnosis is often grouped in four levels (Rytter, 1993), starting by the damage detection. In this level the objective is to know whether there are some changes in the structure and if these changes are due to damage. Second level considers the damage localization. Third level is used to define the type of damage and its size. The last level is defined for calculating the Remaining lifetime of the structure. Recently, an extra level is considered which includes intelligent structures with auto-healing (Figure 1).

Figure 1.

Level in SHM

Complete Chapter List

Search this Book: