The deployment cost of the structural health monitoring (SHM) system is the major argument against the more widespread use of the structural health monitoring techniques. Optimization of sensor placement offers an opportunity to reduce the cost of the SHM system without compromising on the quality of the monitoring approach. Several studies in the area of optimization of sensor placement for SHM applications have been undertaken but the approach has been rather application specific. This article is an attempt to present an unbiased state of the art of the work carried out in the area. The article is targeted towards researchers working in the field of structural health monitoring and optimization of sensor placement as well as practising engineers. This article reviews the work in the area of optimization of sensor placement. It first presents the definition of the optimization problem and then describes each step of the optimization. The current state of the art is then classified based on the techniques for which the optimization of sensor placement has been optimized. The article covers vibration-based monitoring, strain monitoring and elastic wave-based monitoring, as in the eyes of the authors these three techniques are most commonly used and accepted in the SHM community. The article later discusses the different optimization algorithms that have been applied in the literature. The article highlights the different pitfalls of the optimization algorithms and the countermeasures different researchers have proposed to overcome the known shortcomings. In the later section, the multi-objective optimization or the problem definition, keeping in mind the structural as well as executional demands, is discussed. A section has also been developed to showcase the use of optimization of sensor placement techniques’ data fusion–based systems.

No abstract

The joints between structural elements should ensure safe usage of the structure. One of the joining method is based on adhesive bonding. However, adhesive bonding has not replaced riveting yet. Rivets are still present even in newest composite aircraft AIRBUS 350. The reliability of the adhesive bonding limits the use of adhesive bonding for primary aircraft structures and there is a search for new non-destructive testing tools allowing to (1) assessment of the surfaces before bonding and (2) assessment of the adhesive bond. The performance of adhesive bonds depends on the physicochemical properties of the bonded surfaces. The contamination leading to weak bonds may have various origin and be caused by contamination (moisture, release agent, hydraulic fluid and fuel) or poor curing of adhesive. In this work, the research is focused on the development of the method for assessment of the adhesive bonds. Bonded carbon fibrereinforced polymer samples were considered. Electromechanical impedance technique was proposed. The technique is based on electrical impedance measurements of a piezoelectric transducer attached to the investigated structure. The piezoelectric effect causes the electrical response of a piezoelectric transducer to be related to mechanical response of the structure. The indexes for comparison of the conductance spectra were proposed. Three different cases of possible weak bonds were selected for the investigation. The same cases were investigated by destructive methods by other authors. Such approach allows for direct comparison of the obtained results. It was shown that the proposed method allows for clear separation of weak bond cases from the cases for other samples and free sensors. In terms of weak bond assessment, the frequency change with weak bond level (contamination and level of poor curing) was observed. The obtained results are promising and encourage to future research.

This article deals with damage detection process under varying temperature. Carbon fibre–reinforced polymer samples are investigated using electromechanical impedance method. In the article, influence of changing temperature on resistance in electromechanical impedance is investigated. Authors propose new approach for compensation of temperature influence on damage detection. Damage detection is based on root mean square deviation index. Due to strong damping of utilized composite material, low-frequency range is utilized in this research. Real part of electromechanical impedance is measured for frequency band 1–20 kHz. Damage is in the form of artificially made delamination with different sizes. Authors also discuss the problem of influence of structure’s boundary condition on low-frequency measurements. In the research, scanning laser vibrometry for guided wave propagation method is utilized for visualization of the introduced delamination.

Guided waves (GW) allow fast inspection of a large area and hence have attracted research interest from the structural health monitoring (SHM) community. Thus, GW-based SHM is ideal for thin structures such as plates, pipes, etc., and is finding applications in several fields like aerospace, automotive, wind energy, etc. The GW propagate along the surface of the sample and get reflected from discontinuities in the structure in the form of boundaries and damage. Through proper signal processing of the reflected waves based on their time of arrival, the damage can be detected and isolated. For complex structures, a higher number of sensors may be required, which increases the cost of the equipment, as well as the mass. Thus, there is an effort to reduce the number of sensors without compromising the quality of the monitoring achieved. It is of utmost importance that the entire structure can be investigated. Hence, it is necessary to optimize the locations of the sensors in order to maximize the coverage while limiting the number of sensors used. A genetic algorithm (GA)-based optimization strategy was proposed by the authors for use in a simple aluminum plate. This paper extends the optimization methodology for other shape plates and presents experimental, analytical, and numerical studies. The sensitivity studies have been carried out by changing the relative weights of the application demands and presented in the form of a Pareto front. The Pareto front allows comparison of the relative importance of the different application demands, and an appropriate choice can be made based on the information provided.

No abstract

A method for damage localization based on the phased array idea has been developed. Four arrays of transducers are used to perform a beam-forming procedure. Each array consists of nine transducers placed along a line, which are able to excite and register elastic waves. The A 0 Lamb wave mode has been chosen for the localization method. The arrays are placed in such a way that the angular difference between them is 458 and the rotation point is the middle transducer, which is common for all the arrays. The idea has been tested on a square aluminium plate modeled by the Spectral Element Method. Two types of damage were considered, namely distributed damage, which was modeled as stiffness reduction, and cracks, modeled as separation of nodes between selected spectral elements. The plate is excited by a wave packet. The whole array system is placed in the middle of the plate. Each linear phased array in the system acts independently and produces maps of a scanned field based on the beam-forming procedure. These maps are made of time signals (transferred to space domain) that represent the difference between the damaged plate signals and those from the intact plate. An algorithm was developed to join all four maps. The final map is modified by proposed signal processing algorithm to indicate the damaged area of the plate more precisely. The problem for damage localization was investigated and exemplary maps confirming the effectiveness of the proposed system were obtained. It was also shown that the response of the introduced configuration removes the ambiguity of damage localization normally present when a linear phased array is utilized. The investigation is based exclusively on numerical data.

In this paper results of investigation on concentrated piezoelectric networks with different configurations are presented. They were used for elastic wave generation and acquisition. The elastic wave propagation phenomenon was used for damage localization in thin aluminium panels. This approach utilized the fact that any discontinuities existing in structural elements cause local changes of physical material properties which affect elastic wave propagation. Elastic waves were excited and received using piezoelectric transducer networks with different element arrangements. The method of transducer placement and the number of piezoelectric elements used had an influence on the accuracy of the damage localization algorithm. Obviously, the more elements there were, the more data had to be processed. After the acquisition process signal processing was conducted in order to create damage influence maps. These maps presents elastic wave energy connected with reflection from discontinuities. In order to create such a map a computer program was developed that assigns a mesh of points to the panel surface. At each point the energy of elastic wave reflection was calculated. This energy was extracted from the acquired signals. This paper summarizes an extensive experimental investigation that included three damage scenarios and twelve transducer configurations.